Advertisement

Serum neurofilament levels in patients with multiple sclerosis: A comparison of SIMOA and high sensitivity ELISA assays and contributing factors to ELISA levels

Open AccessPublished:September 10, 2022DOI:https://doi.org/10.1016/j.msard.2022.104177

      Abstract

      Background

      Serum neurofilaments (sNfs), especially the most investigated serum neurofilament light chain (sNfL), are promising biomarkers in multiple sclerosis (MS). However, their clinical utility is still limited, given the availability and costs of accessible analytical methods. The gold standard for the detection of sNfs is represented by the single molecule arrays (SIMOA). Recently, a high sensitivity enzyme-linked immunosorbent assay (hsELISA) has also been introduced. The objective of the study was to compare both assays for the determination of sNfL and neurofilament heavy chain (sNfH) concentrations in a defined MS cohort. The second objective was to identify contributing factors to sNfs concentrations determined by hsELISA.

      Methods

      Serum samples were collected from MS patients attending the MS Centre, University Hospital Ostrava, Czech Republic. The levels of sNfs were detected using SIMOA and hsELISA assays.

      Results

      The Spearman's rank correlation coefficient between the sNfL SIMOA and sNfL hsELISA and between the sNfH SIMOA and sNfH hsELISA was moderate rs= 0.543 (p = 0.001) and rs= 0.583 (p = 0.001), respectively. The Passing-Bablok regression analysis demonstrated bias between both methods. Equally significant bias between the methods was confirmed by the Bland-Altman plots. Furthermore, confounding factors affecting the sNfL levels were glomerular filtration rate (eGFR; 95% CI -2.34 to -0.04) and sex (95% CI -2.38 to -0.10). The sNfH levels were affected by age (95% CI 0.01 to 0.07), eGFR (95% CI -2.45 to -0.02), body mass index (BMI; 95% CI -0.31 to -0.05), and blood volume (95% CI 0.69 to 3.35).

      Conclusion

      This analytical study showed significant differences between hsELISA and SIMOA methods, especially for the sNfH concentrations. We identified confounding factors for sNfs levels determined by hsELISA. The sNfs levels were influenced by renal function and sex, whilst sNfH levels were affected by age, BMI, and total blood volume.

      Keywords

      1. Introduction

      The neurofilaments (Nfs) are the main elements of neuroaxonal cytoskeletal proteins (class IV intermediate protein family). They are composed of four subunits, namely the Nf light (NfL), Nf medium, Nf heavy (NfH) chain, and alfa-internexin (
      • Kušnierová P.
      • Zeman D.
      • Hradílek P.
      • Čábal M.
      • Zapletalová O.
      Neurofilament levels in patients with neurological diseases: a comparison of neurofilament light and heavy chain levels.
      ). During any neuroaxonal damage, Nfs are released into the cerebrospinal fluid (CSF) and they can also be detected at lower concentrations in the peripheral blood (
      • Bittner S.
      • Oh J.
      • Havrdová E.K.
      • Tintoré M.
      • Zipp F.
      The potential of serum neurofilament as biomarker for multiple sclerosis.
      ). Consequently, increased levels of Nfs have been described in a spectrum of neurodegenerative diseases, including multiple sclerosis (MS) (
      • Bridel C.
      • Van Wieringen W.N.
      • Zetterberg H.
      • Tijms B.M.
      • Teunissen C.E.
      • Alvarez-Cermeño J.C.
      • Andreasson U.
      • Axelsson M.
      • Bäckström D.C.
      • Bartos A.
      • Bjerke M.
      • Blennow K.
      • Boxer A.
      • Brundin L.
      • Burman J.
      • Christensen T.
      • Fialová L.
      • Forsgren L.
      • Frederiksen J.L.
      • Gisslén M.
      • Gray E.
      • Gunnarsson M.
      • Hall S.
      • Hansson O.
      • Herbert M.K.
      • Jakobsson J.
      • Jessen-Krut J.
      • Janelidze S.
      • Johannsson G.
      • Jonsson M.
      • Kappos L.
      • Khademi M.
      • Khalil M.
      • Kuhle J.
      • Landén M.
      • Leinonen V.
      • Logroscino G.
      • Lu C.H.
      • Lycke J.
      • Magdalinou N.K.
      • Malaspina A.
      • Mattsson N.
      • Meeter L.H.
      • Mehta S.R.
      • Modvig S.
      • Olsson T.
      • Paterson R.W.
      • Pérez-Santiago J.
      • Piehl F.
      • Pijnenburg Y.A.L.
      • Pyykkö O.T.
      • Ragnarsson O.
      • Rojas J.C.
      • Romme Christensen J.
      • Sandberg L.
      • Scherling C.S.
      • Schott J.M.
      • Sellebjerg F.T.
      • Simone I.L.
      • Skillbäck T.
      • Stilund M.
      • Sundström P.
      • Svenningsson A.
      • Tortelli R.
      • Tortorella C.
      • Trentini A.
      • Troiano M.
      • Turner M.R.
      • Van Swieten J.C.
      • Vågberg M.
      • Verbeek M.M.
      • Villar L.M.
      • Visser P.J.
      • Wallin A.
      • Weiss A.
      • Wikkelsø C.
      • Wild E.J.
      Diagnostic value of cerebrospinal fluid neurofilament light protein in neurology: a systematic review and meta-analysis.
      ). Serum Nfs (sNfs), especially the most investigated serum NfL (sNfL), are one of the most promising biomarkers in MS. Serum NfL levels reflect inflammatory disease activity, and are an easily accessible prognostic marker and indicator of treatment response (
      • Bittner S.
      • Oh J.
      • Havrdová E.K.
      • Tintoré M.
      • Zipp F.
      The potential of serum neurofilament as biomarker for multiple sclerosis.
      ). However, their clinical utility is still limited, given the availability and costs of accessible analytical methods. The gold standard for the sNfs detection is represented by the fourth-generation immunoassay, single molecule arrays (SIMOA). The SIMOA is an ultrasensitive digital immunoassay for the quantitative determination of protein biomarkers with high precision and performance, but it is designated for research use only (RUO). Recently, a less-expensive high sensitivity enzyme-linked immunosorbent assay (hsELISA) was introduced to determinate the sNfL with a similar detection limit for sNfL quantification as the SIMOA (
      • Fialová L.
      • Nosková L.
      • Kalousová M.
      • Zima T.
      • Uher T.
      • Bartoš A.
      Analytical and pre-analytical aspects of neurofilament light chain determination in biological fluids.
      ). Although both methods are intended for RUO, ELISA is more accessible in most laboratories. The hsELISA represents the only laboratory method intended for in vitro diagnostics of serum NfH (sNfH).
      The main objective of this study was to compare both assays for the determination of sNfs concentrations in a defined MS cohort. The second objective was to identify factors contributing to sNfs concentrations determined by hsELISA.

      2. Material and methods

      2.1 Patients and data collection

      Serum samples were collected from MS patients attending the MS Centre, University Hospital Ostrava, Ostrava, Czech Republic, who were included into a single-centre prospective cohort study (”Dynamics of serum biomarkers in monitoring the effectiveness of multiple sclerosis treatment with DMD preparations”). The inclusion criteria were: (i) clinically isolated syndrome (CIS) or MS according to the latest 2017 revision of the McDonald criteria (
      • Thompson A.J.
      • Banwell B.L.
      • Barkhof F.
      • Carroll W.M.
      • Coetzee T.
      • Comi G.
      • Correale J.
      • Fazekas F.
      • Filippi M.
      • Freedman M.S.
      • Fujihara K.
      • Galetta S.L.
      • Hartung H.P.
      • Kappos L.
      • Lublin F.D.
      • Marrie R.A.
      • Miller A.E.
      • Miller D.H.
      • Montalban X.
      • Mowry E.M.
      • Sorensen P.S.
      • Tintoré M.
      • Traboulsee A.L.
      • Trojano M.
      • Uitdehaag B.M.J.
      • Vukusic S.
      • Waubant E.
      • Weinshenker B.G.
      • Reingold S.C.
      • Cohen J.A.
      Diagnosis of multiple sclerosis: 2017 revisions of the McDonald criteria.
      ), (ii) indication for starting or change of disease modifying treatments (DMTs), (iii) age > 18 years, and (iv) signed informed consent. The exclusion criteria were: (i) not confirmed diagnosis of CIS or MS, (ii) age < 18 years. The baseline data of the prospective study were used for the study of contributing factors. Random samples from the study participants were used for the comparison of assays, the selection was made blind to the performing laboratory worker.
      Demographic data (age and sex), self-reported weight and height, and self-reported cardiological comorbidities were collected at the baseline visit. Blood volume was calculated using the Nadler et al. formula (
      • Nadler S.B.
      • Hidalgo J.U.
      • Bloch T.
      Prediction of blood volume in normal human adults.
      ). Body mass index (BMI) was calculated from the patient`s self-reported height and weight.
      Ongoing DMTs were categorised into moderately effective DMTs (all interferons, dimethyl fumarate, glatiramer acetate and teriflunomide) and highly effective DMTs (alemtuzumab, cladribine, fingolimod, natalizumab, ofatumumab and ocrelizumab).

      2.2 Samples

      Serum samples were collected into a serum gel with clotting activator tube (Sarstedt). After delivery to the laboratory, the serum samples were centrifuged at 2500 × g for 6 min at 4°C, then aliquoted into at least three vials (0.3 mL per vial) and stored at −70°C until the analysis.

      2.3 Analytical methods

      The concentrations of sNfL and serum NfH (sNfH) were determined by SIMOA (NF-light™ Advantage Kit HD-1/HD-X, REF 103186, RUO; pNF-heavy Discovery Kit HD-1/HD-X, REF 102669, RUO; Quanterix Corporation) and hsELISA assays (Nf-light serum ELISA, REF. 30210101, RUO, UmanDiagnostics AB; Neurofilament (pNf‐H) ‐ high sensitive ELISA, REF EQ6562‐9601, IVD, Euroimmun AG). The kit manufacturers stated that the limits of detection were 0.038 ng/L for sNfL SIMOA, 0.663 ng/L for sNfH SIMOA, 0.4 ng/L for sNfL hsELISA and 1.7 ng/L for sNfH hsELISA, respectively. All samples were analysed in duplicates. Measurements of sNfL and sNfH SIMOA were performed using a 4X diluted serum. Intra-assay coefficients of variation for sNfL were 0.0 to 7.7% and the inter-assay coefficients of variation were 0.0 to 6.4%, and for sNfH 2.7 to 10.1% and 4.2 to 15.6%, respectively. For sNfL hsELISA, the 4X diluted serum was used, whereas for sNfH hsELISA undiluted serum samples were used. The intra-assay coefficients of variation for sNfL hsELISA were 4.3 to 6.6%, and 4.0 to 6.0% for sNfH hsELISA. The inter-assay coefficients of variation for sNfL hsELISA and sNfH hsELISA were 5.7 to 10.0%, and 4.7 to 10.6%, respectively.
      Creatinine was measured by enzymatic assay on Atellica® CH Analyzer with Atellica CH Enzymatic Creatinine_2 (ECre_2) kit, Siemens Healthcare Diagnostics Inc.) Estimated glomerular filtration rate (eGFR) was calculated from serum creatinine according to CKD-EPI Equation (
      • Levin A.
      • Stevens P.E.
      • Bilous R.W.
      • Coresh J.
      • De Francisco A.L.M.
      • De Jong P.E.
      • Griffith K.E.
      • Hemmelgarn B.R.
      • Iseki K.
      • Lamb E.J.
      • Levey A.S.
      • Riella M.C.
      • Shlipak M.G.
      • Wang H.
      • White C.T.
      • Winearls C.G.
      Kidney disease: improving global outcomes (KDIGO) CKD work group. KDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease.
      ).

      2.4 Statistical methods

      R software version 4.2.0 was used for statistical analysis. Descriptive statistics were used to outline our data. The categorical variables were evaluated using frequency tables. The continuous and integer variables were reported as medians and interquartile ranges (IQR) because of a non-normal distribution.
      Firstly, a comparison of analytical methods was performed. The results of bias or agreement with the declared values of commercial control materials were assessed using a one-sample t-test. Secondly, a comparison of analytical methods was performed based on our MS samples. The normality of the sNfL and sNfH concentration was evaluated by plotting a histogram. The normality was rejected; therefore, nonparametric tests were used. Spearman's rank correlation coefficient and the Passing-Bablok regression analysis assessed the relationship between both assays. The Spearman's rank correlation coefficient was interpreted according to
      • Prion S.
      • Haerling K.A.
      Making sense of methods and measurement: Spearman-Rho Ranked-Order Correlation Coefficient.
      ). For the comparison of systematic differences, a Bland-Altman plot was performed.
      The third step was to evaluate the influence of factors (age, total blood volume, eGFR, reported cardiological comorbidities, and sex) on sNfs levels. The normality of sNfL and sNfH levels was evaluated by plotting a histogram. For a non-normal distribution, the sNfs values were logarithmically transformed. Simple and multiple linear regression were used to determine the association. The significance level was set at 5%.

      2.5 Ethics approval

      Informed consent was obtained from all patients at the University Hospital Ostrava who were included in the study. The study was approved by the Ethics Committee of the University Hospital Ostrava as a part of the project “Dynamics of serum biomarkers in monitoring the effectiveness of multiple sclerosis treatment with DMD preparations” (reference number 338/2020).

      3. Results

      3.1 Assessment of precision and trueness of individual analytical methods using commercial controls

      ELISA methods obtained the lowest coefficient of variation values when evaluating accuracy under repeatability conditions (Table 1). The results of bias or agreement with the declared values of commercial control materials assessed using a one-sample t-test of both methods were comparable (Table 1). When comparing the obtained bias value with the maximum permissible bias value calculated on the basis of the biological variability of sNfL (< 8.9%) (
      • Hviid C.
      • Madsen A.T.
      • Winther-Larsen A.
      Biological variation of serum neurofilament light chain.
      ) a comparable bias value of both methods was found. Due to the lack of knowledge of biological variability of sNfH, it was not possible to comment on the bias value of this test.
      Table 1Assessment of precision and trueness of individual analytical methods using commercial controls.
      AnalyteLevel 1Level 2
      Mean (ng/L)SDCVBias (%)P valueMean (ng/L)SDCVBias (%)P value
      (%)(%)
      sNfL hsELISA16.00.955.965.630.10---
      sNfL hsSIMOA3.240.4213.0-4.420.4716519.511.92.810.60
      sNfH hsELISA33.92.005.90-3.030.2991.38.679.49-0.780.87
      sNfH hsSIMOA13812.79.176.310.23256527310.72.600.62
      Abbreviations: sNfL: serum neurofilament light chain; sNfH: serum neurofilament heavy chain; SD: standard deviation; CV: coefficient of variation.

      3.2 The comparison of SIMOA and high sensitivity ELISA assays in the MS cohort

      A total of 32 patients’ samples were used for the comparison of both assays. The median sNfL concentration was 11.65 ng/L (IQR 9.48 – 16.03) using SIMOA and 8.40 ng/L (IQR 6.30 – 11.28) using hsELISA. The median sNfH concentration was 34.85 ng/L (IQR 9.28 – 53.28) using SIMOA and 5.96 ng/L (IQR 1.70 – 26.0) using hsELISA. The Spearman's rank correlation coefficient between the sNfL SIMOA and sNfL hsELISA, and between the sNfH SIMOA and sNfH hsELISA was moderate rs= 0.543 (p = 0.001) and rs= 0.583 (p = 0.001), respectively. The Passing-Bablok regression analysis demonstrated bias between both methods (Fig. 1). Equally significant bias between the methods was confirmed by the Bland-Altman plots (Fig. 2). On average, SIMOA measured 3.69 ng/L (29.8%) higher values of sNfL, and 20.94 ng/L (87.1%) higher value of sNfH than hsELISA. The differences in measured concentrations were also clinically significant, i.e. greater than the allowed error for determination (25%).
      Fig 1
      Fig. 1The Passing-Bablok analysis of (A) SIMOA sNfL vs hsELISA sNfL and (B) SIMOA sNfH vs hsELISA sNfH
      Fig 2
      Fig. 2Bland-Altman plot with limits of agreement of (A) SIMOA sNfL vs hsELISA sNfL and (B) SIMOA sNfH vs hsELISA sNfH in units (ng/L).
      Legend: The shaded blue represents the bias with 95% confidence interval, green represents upper limit of agreement with 95% confidence interval, and red lower limit of agreement with 95% confidence interval.

      3.3 High sensitivity ELISA sNfL levels and contributing factors

      The median sNfL concentration (n = 60) was 8.85 ng/L (IQR 6.10 – 16.57 ng/L). The demographic characteristics of the patients are presented in Table 2. In the simple regression model, there was no statistically significant association between log sNfL levels and age, sex, BMI, blood volume, reported cardiological comorbidities, and eGFR. There were lower log sNfL levels in women in the multivariate model (95% CI -2.38 to -0.10, p = 0.033). There was a significant association between eGFR and sNfL in the multiple regression model with an average decrease of -1.19 ng/L (95% CI -2.34 to -0.04, p = 0.043) in log sNfL levels per 1 mL/s increase in eGFR (Table 3, Fig. 3).
      Table 2Characteristics of the study participants.
      sNfLsNfH
      Number of samples6069
      Age (median, IQR)36.5 (27.6–46.5)37.0 (28.1 – 45.4)
      Women (n, %)42 (75.0)49 (71.0)
      BMI (median, IQR)24.2 (22.0– 27.5)23.7 (21.6–26.8)
      eGFR (median, IQR)1.75 (1.59–1.90)1.76 (1.60 – 1.93)
      Blood volume in litres (median, IQR)3.20 (2.91 – 3.77)3.18 (2.88 – 3.75)
      Ongoing DMTs (n, %)
      Moderately effective DMTs48 (80.0)54 (78.3)
      Highly effective DMTs0 (0)3 (4.3)
      Time on DMTs in days (median, IQR)28 (23 – 30)28 (23 – 33)
      Cardiological comorbidities (n, %)8 (13.3)8 (11.6)
      Abbreviations: n: number; IQR: interquartile range; sNfL: serum neurofilament light chain; sNfH: serum neurofilament heavy chain; BMI: body mass index; eGFR: estimated glomerular filtration rate; DMTs: disease modifying treatments.
      Table 3Associations between hsELISA neurofilament levels and contributing factors.
      OutcomeNPredictorSimple modelMultiple model
      ß (95% CI)R2P valueß (95% CI)R2P value
      sNfL60Age-0.003 (-0.02 to 0.02)0.000.816-0.02 (-0.04 to 0.01)0.150.261
      60Women-0.33 (-0.87 to 0.21)0.030.227-1.24 (-2.38 to -0.10)0.150.033
      60BMI-0.01 (-0.06 to 0.05)0.000.8500.09 (-0.03 to 0.20)0.150.121
      60Blood volume0.04 (-0.37 to 0.46)0.000.846-1.12 (-2.34 to 0.11)0.150.073
      60Cardiological comorbidities
      Included arterial hypertension, chronic venous insufficiency, congenital valve defect, and deep vein thrombosis and pulmonary embolism.
      -0.35 (-1.08 to 0.38)0.020.340-0.37 (-1.14 to 0.40)0.150.341
      60eGFR-0.83 (-1.87 to 0.22)0.040.118-1.19 (-2.34 to -0.04)0.150.043
      sNfH69Age0.04 (0.02 to 0.07)0.140.0020.03 (0.01 to 0.06)0.300.020
      69Women-0.25 (-0.90 to 0.40)0.010.4481.33 (0.13 to 2.53)0.300.030
      69BMI0.04 (-0.03 to 0.10)0.020.236-0.16 (-0.29 to -0.03)0.300.018
      69Blood volume0.48 (-0.01 to 0.96)0.050.0551.83 (0.51 to 3.15)0.300.008
      69Cardiological comorbidities
      Included arterial hypertension, chronic venous insufficiency, congenital valve defect, and deep vein thrombosis and pulmonary embolism.
      1.94 (-0.25 to 1.57)0.030.1540.47 (-0.40 to 1.33)0.300.287
      69eGFR-1.95 (-3.12 to -0.78)0.140.001-1.23 (-2.45 to -0.02)0.300.046
      Abbreviations: N: number; ß: beta coefficient; 95% CI: 95% confidence interval; R2: R- squared; sNfL: serum neurofilament light chain; sNfH: serum neurofilament heavy chain; BMI: body mass index; eGFR: estimated glomerular filtration rate.
      Included arterial hypertension, chronic venous insufficiency, congenital valve defect, and deep vein thrombosis and pulmonary embolism.
      Fig 3
      Fig. 3Coefficient plots with 95% confidence interval from the multiple linear regression models (A) for sNfL levels and (B) for sNfH levels.
      Abbreviations: BMI: body mass index; eGFR: estimated glomerular filtration rate.

      3.4 High sensitivity ELISA sNfH levels and contributing factors

      The median sNfH concentration (n = 69) was 6.45 ng/L (IQR 2.45 – 18.09 ng/L). The demographic characteristics of the patients are presented in Table 2. No association was found between sex, BMI, blood volume, reported cardiological comorbidities and log sNfH levels in the univariate model. However, a significant association between log sNfH and BMI was found in the multiple regression model (95% CI -0.31 to -0.05, p = 0.009), and log sNfH and blood volume with an average increase of 2.02 ng/L (95% CI 0.69 to 3.35, p = 0.004) in log sNfH per litre increase in blood volume. Similarly, the female sex was associated with higher log sNfH in the multiple regression model (95% 0.26 to 2.66, p = 0.018). A significant positive association between log sNfH and age was found in the simple and multiple regression model (95% CI 0.01 to 0.07, p = 0.014). There was a negative association between sNfH and eGFR (95% CI -2.45 to -0.02, p = 0.046) in both models.

      4. Discussion

      In this analytical study, we compared the SIMOA and hsELISA assays for the measurement of sNfs concentration for the first time in existing literature. The results were rather discouraging, showing large differences between these methods, especially for sNfH. For sNfL concentrations, the differences between hsELISA and SIMOA were more acceptable than for ELISA (
      • Kuhle J.
      • Barro C.
      • Andreasson U.
      • Derfuss T.
      • Lindberg R.
      • Sandelius Å.
      • Liman V.
      • Norgren N.
      • Blennow K.
      • Zetterberg H.
      Comparison of three analytical platforms for quantification of the neurofilament light chain in blood samples: ELISA, electrochemiluminescence immunoassay and Simoa.
      ). Although SIMOA was proposed as a potential gold standard for blood Nfs determination, the high costs of the method prevent it being used worldwide. Hence, the improvements in ELISA methods which are by far less expensive and easy to perform in most laboratories, are to be welcomed.
      Several studies have addressed the issue of potential confounders on blood Nfs concentrations measured by SIMOA assays. Our study demonstrated a negative association between sNfs measured by hsELISA and eGFR, which was previously observed in non-MS aging populations (
      • Akamine S.
      • Marutani N.
      • Kanayama D.
      • Gotoh S.
      • Maruyama R.
      • Yanagida K.
      • Sakagami Y.
      • Mori K.
      • Adachi H.
      • Kozawa J.
      • Maeda N.
      • Otsuki M.
      • Matsuoka T.
      • Iwahashi H.
      • Shimomura I.
      • Ikeda M.
      • Kudo T.
      Renal function is associated with blood neurofilament light chain level in older adults.
      ;
      • Ladang A.
      • Kovacs S.
      • Lengelé L.
      • Locquet M.
      • Reginster J.Y.
      • Bruyère O.
      • Cavalier E.
      Neurofilament light chain concentration in an aging population.
      ). The reason for such association remains unclear but may be related to the presence of lower molecular weight fragments of sNfs (recognised by the antibodies used in sNfs assays) that are filtered in the glomerulus (
      • Ladang A.
      • Kovacs S.
      • Lengelé L.
      • Locquet M.
      • Reginster J.Y.
      • Bruyère O.
      • Cavalier E.
      Neurofilament light chain concentration in an aging population.
      ). Estimation of molecular weights of Nfs proteins in blood by SDS electrophoresis could possibly give an answer. With further advances in mass spectrometry and improvement in assay sensitivity, determination of molecular weights of Nfs proteins in blood could also become available in the future.
      • Akamine S.
      • Marutani N.
      • Kanayama D.
      • Gotoh S.
      • Maruyama R.
      • Yanagida K.
      • Sakagami Y.
      • Mori K.
      • Adachi H.
      • Kozawa J.
      • Maeda N.
      • Otsuki M.
      • Matsuoka T.
      • Iwahashi H.
      • Shimomura I.
      • Ikeda M.
      • Kudo T.
      Renal function is associated with blood neurofilament light chain level in older adults.
      also speculated about an indirect association between decreased kidney function and damage of the central nervous system. This theory was based on low levels of erythropoietin and vitamin D (both of which were reported to have neuroprotective effects) in patients with impaired renal function (
      • Akamine S.
      • Marutani N.
      • Kanayama D.
      • Gotoh S.
      • Maruyama R.
      • Yanagida K.
      • Sakagami Y.
      • Mori K.
      • Adachi H.
      • Kozawa J.
      • Maeda N.
      • Otsuki M.
      • Matsuoka T.
      • Iwahashi H.
      • Shimomura I.
      • Ikeda M.
      • Kudo T.
      Renal function is associated with blood neurofilament light chain level in older adults.
      ).
      Other factors influencing the sNfs concentrations are less clear and the body of evidence is inconsistent. We have found a positive association of sNfH with age. CSF NfH were correlated with age in previous studies in both control groups and MS patients (
      • Kuhle J.
      • Leppert D.
      • Petzold A.
      • Regeniter A.
      • Schindler C.
      • Mehling M.
      • Anthony D.C.
      • Kappos L.
      • Lindberg R.L.P.
      Neurofilament heavy chain in CSF correlates with relapses and disability in multiple sclerosis.
      ;
      • Teunissen C.E.
      • Iacobaeus E.
      • Khademi M.
      • Brundin L.
      • Norgren N.
      • Koel-Simmelink M.J.A.
      • Schepens M.
      • Bouwman F.
      • Twaalfhoven H.A.M.
      • Blom H.J.
      • Jakobs C.
      • Dijkstra C.D.
      Combination of CSF N-acetylaspartate and neurofilaments in multiple sclerosis.
      ). Age was not an independent predictive factor for sNfH levels in one published study (
      • Gresle M.
      • Liu Y.
      • Dagley L.F.
      • Haartsen J.
      • Pearson F.
      • Purcell A.W.
      • Laverick L.
      • Petzold A.
      • Lucas R.M.
      • Van Der Walt A.
      • Prime H.
      • Morris D.R.
      • Taylor B.V.
      • Shaw G.
      • Butzkueven H.
      Serum phosphorylated neurofilament-heavy chain levels in multiple sclerosis patients.
      ). While we found a negative association between sNfH and BMI, we were not able to confirm a similar association for sNfL, at variance with previous studies (
      • Manouchehrinia A.
      • Piehl F.
      • Hillert J.
      • Kuhle J.
      • Alfredsson L.
      • Olsson T.
      • Kockum I.
      Confounding effect of blood volume and body mass index on blood neurofilament light chain levels.
      ;
      • Nilsson I.A.K.
      • Millischer V.
      • Karrenbauer V.D.
      • Juréus A.
      • Salehi A.M.
      • Norring C.
      • von Hausswolff-Juhlin Y.
      • Schalling M.
      • Blennow K.
      • Bulik C.M.
      • Zetterberg H.
      • Landén M.
      Plasma neurofilament light chain concentration is increased in anorexia nervosa.
      ). Sex differences were inconsistent for sNfL and sNfH in our study and may reflect different disease characteristics since no sex differences were observed for sNfL apart from amyotrophic lateral sclerosis (
      • Barro C.
      • Chitnis T.
      • Weiner H.L.
      Blood neurofilament light: a critical review of its application to neurologic disease.
      ;
      • Fialová L.
      • Nosková L.
      • Kalousová M.
      • Zima T.
      • Uher T.
      • Bartoš A.
      Analytical and pre-analytical aspects of neurofilament light chain determination in biological fluids.
      ), where higher concentrations in female patients were found (
      • Lu C.H.
      • Macdonald-Wallis P.
      • Gray E.
      • Pearce N.
      • Petzold A.
      • Norgren N.
      • Giovannoni G.
      • Fratta P.
      • Sidle K.
      • Fish M.
      • Orrell R.
      • Howard R.
      • Talbot K.
      • Greensmith L.
      • Kuhle J.
      • Turner M.R.
      • Malaspina A.
      Neurofilament light chain. A prognostic biomarker in amyotrophic lateral sclerosis.
      ;
      • Thouvenot E.
      • Demattei C.
      • Lehmann S.
      • Maceski-Maleska A.
      • Hirtz C.
      • Juntas-Morales R.
      • Pageot N.
      • Esselin F.
      • Alphandéry S.
      • Vincent T.
      • Camu W.
      Serum neurofilament light chain at time of diagnosis is an independent prognostic factor of survival in amyotrophic lateral sclerosis.
      ). It was explained by the higher severity of the disease in female versus male patients (
      • Barro C.
      • Chitnis T.
      • Weiner H.L.
      Blood neurofilament light: a critical review of its application to neurologic disease.
      ). A positive association between estimated blood volume and sNfH concentration is at variance with results obtained for NfL by
      • Manouchehrinia A.
      • Piehl F.
      • Hillert J.
      • Kuhle J.
      • Alfredsson L.
      • Olsson T.
      • Kockum I.
      Confounding effect of blood volume and body mass index on blood neurofilament light chain levels.
      . Deeper understanding of the kinetics of Nfs proteins is needed to explain the observed associations.
      Our study has several limitations. The modest limitation is a limited sample size and missing control group of healthy controls as it was not originally our intention to compare these analytical methods as part of the prospective study. Other limitations are the patient self-reported comorbidities, weight and height, which may potentially bias the results.

      5. Conclusions

      In this analytical study of patients with confirmed MS, we compared the SIMOA and hsELISA assays in the determination of sNfs. Our results showed significant differences between these methods, especially for sNfH concentrations. Furthermore, we identified confounding factors for sNfs levels determined by hsELISA. The sNfs levels were influenced by renal function and sex, sNfH by age, BMI, and total blood volume. Further studies with a larger sample size and control group are warranted.

      Funding information

      The study was supported in part by an institutional grant from the Ministry of Health, Czech Republic (12/RVO-FNOs/2020).

      Data Availability Statement

      De-identified data will be shared by the corresponding author with any qualified investigator by request.

      Study funding

      The study was supported in part by an institutional grant from the Ministry of Health , Czech Republic ( 12/RVO-FNOs/2020 ).

      CRediT authorship contribution statement

      K. Zondra Revendova: Conceptualization, Resources, Methodology, Data curation, Formal analysis, Funding acquisition, Investigation, Project administration, Writing – original draft, Writing – review & editing. D. Zeman: Conceptualization, Writing – original draft, Validation, Resources, Methodology, Funding acquisition, Investigation, Writing – review & editing. R. Bunganic: Conceptualization, Resources, Funding acquisition, Methodology, Investigation. K. Karasova: Investigation, Writing – review & editing. O. Volny: Methodology, Writing – original draft, Writing – review & editing. M. Bar: Supervision, Validation, Resources, Writing – review & editing. P. Kusnierova: Conceptualization, Supervision, Methodology, Data curation, Validation, Formal analysis, Funding acquisition, Investigation, Project administration, Resources, Writing – original draft, Writing – review & editing.

      Declaration of Competing Interest

      Authors have no conflicts of interest to declare.

      Acknowledgments

      We would like to thank all the patients and staff contributing to data collection. The authors also thank Frantisek Vsiansky for statistic consultations, and Marcela Ely, PhD for proofreading.

      References

        • Akamine S.
        • Marutani N.
        • Kanayama D.
        • Gotoh S.
        • Maruyama R.
        • Yanagida K.
        • Sakagami Y.
        • Mori K.
        • Adachi H.
        • Kozawa J.
        • Maeda N.
        • Otsuki M.
        • Matsuoka T.
        • Iwahashi H.
        • Shimomura I.
        • Ikeda M.
        • Kudo T.
        Renal function is associated with blood neurofilament light chain level in older adults.
        Sci. Rep. 2020; 10https://doi.org/10.1038/s41598-020-76990-7
        • Barro C.
        • Chitnis T.
        • Weiner H.L.
        Blood neurofilament light: a critical review of its application to neurologic disease.
        Ann. Clin. Transl. Neurol. 2020; https://doi.org/10.1002/acn3.51234
        • Bittner S.
        • Oh J.
        • Havrdová E.K.
        • Tintoré M.
        • Zipp F.
        The potential of serum neurofilament as biomarker for multiple sclerosis.
        Brain. 2021; 144https://doi.org/10.1093/brain/awab241
        • Bridel C.
        • Van Wieringen W.N.
        • Zetterberg H.
        • Tijms B.M.
        • Teunissen C.E.
        • Alvarez-Cermeño J.C.
        • Andreasson U.
        • Axelsson M.
        • Bäckström D.C.
        • Bartos A.
        • Bjerke M.
        • Blennow K.
        • Boxer A.
        • Brundin L.
        • Burman J.
        • Christensen T.
        • Fialová L.
        • Forsgren L.
        • Frederiksen J.L.
        • Gisslén M.
        • Gray E.
        • Gunnarsson M.
        • Hall S.
        • Hansson O.
        • Herbert M.K.
        • Jakobsson J.
        • Jessen-Krut J.
        • Janelidze S.
        • Johannsson G.
        • Jonsson M.
        • Kappos L.
        • Khademi M.
        • Khalil M.
        • Kuhle J.
        • Landén M.
        • Leinonen V.
        • Logroscino G.
        • Lu C.H.
        • Lycke J.
        • Magdalinou N.K.
        • Malaspina A.
        • Mattsson N.
        • Meeter L.H.
        • Mehta S.R.
        • Modvig S.
        • Olsson T.
        • Paterson R.W.
        • Pérez-Santiago J.
        • Piehl F.
        • Pijnenburg Y.A.L.
        • Pyykkö O.T.
        • Ragnarsson O.
        • Rojas J.C.
        • Romme Christensen J.
        • Sandberg L.
        • Scherling C.S.
        • Schott J.M.
        • Sellebjerg F.T.
        • Simone I.L.
        • Skillbäck T.
        • Stilund M.
        • Sundström P.
        • Svenningsson A.
        • Tortelli R.
        • Tortorella C.
        • Trentini A.
        • Troiano M.
        • Turner M.R.
        • Van Swieten J.C.
        • Vågberg M.
        • Verbeek M.M.
        • Villar L.M.
        • Visser P.J.
        • Wallin A.
        • Weiss A.
        • Wikkelsø C.
        • Wild E.J.
        Diagnostic value of cerebrospinal fluid neurofilament light protein in neurology: a systematic review and meta-analysis.
        JAMA Neurol. 2019; 76https://doi.org/10.1001/jamaneurol.2019.1534
        • Fialová L.
        • Nosková L.
        • Kalousová M.
        • Zima T.
        • Uher T.
        • Bartoš A.
        Analytical and pre-analytical aspects of neurofilament light chain determination in biological fluids.
        Ces. Slov. Neurol. Neurochir. 2022; 85: 11-16https://doi.org/10.48095/cccsnn202211
        • Gresle M.
        • Liu Y.
        • Dagley L.F.
        • Haartsen J.
        • Pearson F.
        • Purcell A.W.
        • Laverick L.
        • Petzold A.
        • Lucas R.M.
        • Van Der Walt A.
        • Prime H.
        • Morris D.R.
        • Taylor B.V.
        • Shaw G.
        • Butzkueven H.
        Serum phosphorylated neurofilament-heavy chain levels in multiple sclerosis patients.
        J. Neurol. Neurosurg. Psychiatry. 2014; 85https://doi.org/10.1136/jnnp-2013-306789
        • Hviid C.
        • Madsen A.T.
        • Winther-Larsen A.
        Biological variation of serum neurofilament light chain.
        Clin. Chem. Lab. Med. 2021; 60https://doi.org/10.1515/cclm-2020-1276
        • Kuhle J.
        • Barro C.
        • Andreasson U.
        • Derfuss T.
        • Lindberg R.
        • Sandelius Å.
        • Liman V.
        • Norgren N.
        • Blennow K.
        • Zetterberg H.
        Comparison of three analytical platforms for quantification of the neurofilament light chain in blood samples: ELISA, electrochemiluminescence immunoassay and Simoa.
        Clin. Chem. Lab. Med. 2016; 54https://doi.org/10.1515/cclm-2015-1195
        • Kuhle J.
        • Leppert D.
        • Petzold A.
        • Regeniter A.
        • Schindler C.
        • Mehling M.
        • Anthony D.C.
        • Kappos L.
        • Lindberg R.L.P.
        Neurofilament heavy chain in CSF correlates with relapses and disability in multiple sclerosis.
        Neurology. 2011; 76https://doi.org/10.1212/WNL.0b013e31821432ff
        • Kušnierová P.
        • Zeman D.
        • Hradílek P.
        • Čábal M.
        • Zapletalová O.
        Neurofilament levels in patients with neurological diseases: a comparison of neurofilament light and heavy chain levels.
        J. Clin. Lab. Anal. 2019; 33https://doi.org/10.1002/jcla.22948
        • Ladang A.
        • Kovacs S.
        • Lengelé L.
        • Locquet M.
        • Reginster J.Y.
        • Bruyère O.
        • Cavalier E.
        Neurofilament light chain concentration in an aging population.
        Aging Clin. Exp. Res. 2022; 34https://doi.org/10.1007/s40520-021-02054-z
        • Levin A.
        • Stevens P.E.
        • Bilous R.W.
        • Coresh J.
        • De Francisco A.L.M.
        • De Jong P.E.
        • Griffith K.E.
        • Hemmelgarn B.R.
        • Iseki K.
        • Lamb E.J.
        • Levey A.S.
        • Riella M.C.
        • Shlipak M.G.
        • Wang H.
        • White C.T.
        • Winearls C.G.
        Kidney disease: improving global outcomes (KDIGO) CKD work group. KDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease.
        Kidney Int. Suppl. 2013; https://doi.org/10.1038/kisup.2012.73
        • Lu C.H.
        • Macdonald-Wallis P.
        • Gray E.
        • Pearce N.
        • Petzold A.
        • Norgren N.
        • Giovannoni G.
        • Fratta P.
        • Sidle K.
        • Fish M.
        • Orrell R.
        • Howard R.
        • Talbot K.
        • Greensmith L.
        • Kuhle J.
        • Turner M.R.
        • Malaspina A.
        Neurofilament light chain. A prognostic biomarker in amyotrophic lateral sclerosis.
        Neurology. 2015; 84: 2247-2257
        • Manouchehrinia A.
        • Piehl F.
        • Hillert J.
        • Kuhle J.
        • Alfredsson L.
        • Olsson T.
        • Kockum I.
        Confounding effect of blood volume and body mass index on blood neurofilament light chain levels.
        Ann. Clin. Transl. Neurol. 2020; 7https://doi.org/10.1002/acn3.50972
        • Nadler S.B.
        • Hidalgo J.U.
        • Bloch T.
        Prediction of blood volume in normal human adults.
        Surgery. 1962; 51
        • Nilsson I.A.K.
        • Millischer V.
        • Karrenbauer V.D.
        • Juréus A.
        • Salehi A.M.
        • Norring C.
        • von Hausswolff-Juhlin Y.
        • Schalling M.
        • Blennow K.
        • Bulik C.M.
        • Zetterberg H.
        • Landén M.
        Plasma neurofilament light chain concentration is increased in anorexia nervosa.
        Transl. Psychiatry. 2019; 9https://doi.org/10.1038/s41398-019-0518-2
        • Prion S.
        • Haerling K.A.
        Making sense of methods and measurement: Spearman-Rho Ranked-Order Correlation Coefficient.
        Clin. Simul. Nurs. 2014; 10https://doi.org/10.1016/j.ecns.2014.07.005
        • Teunissen C.E.
        • Iacobaeus E.
        • Khademi M.
        • Brundin L.
        • Norgren N.
        • Koel-Simmelink M.J.A.
        • Schepens M.
        • Bouwman F.
        • Twaalfhoven H.A.M.
        • Blom H.J.
        • Jakobs C.
        • Dijkstra C.D.
        Combination of CSF N-acetylaspartate and neurofilaments in multiple sclerosis.
        Neurology. 2009; 72https://doi.org/10.1212/WNL.0b013e3181a0fe3f
        • Thompson A.J.
        • Banwell B.L.
        • Barkhof F.
        • Carroll W.M.
        • Coetzee T.
        • Comi G.
        • Correale J.
        • Fazekas F.
        • Filippi M.
        • Freedman M.S.
        • Fujihara K.
        • Galetta S.L.
        • Hartung H.P.
        • Kappos L.
        • Lublin F.D.
        • Marrie R.A.
        • Miller A.E.
        • Miller D.H.
        • Montalban X.
        • Mowry E.M.
        • Sorensen P.S.
        • Tintoré M.
        • Traboulsee A.L.
        • Trojano M.
        • Uitdehaag B.M.J.
        • Vukusic S.
        • Waubant E.
        • Weinshenker B.G.
        • Reingold S.C.
        • Cohen J.A.
        Diagnosis of multiple sclerosis: 2017 revisions of the McDonald criteria.
        Lancet Neurol. 2018; https://doi.org/10.1016/S1474-4422(17)30470-2
        • Thouvenot E.
        • Demattei C.
        • Lehmann S.
        • Maceski-Maleska A.
        • Hirtz C.
        • Juntas-Morales R.
        • Pageot N.
        • Esselin F.
        • Alphandéry S.
        • Vincent T.
        • Camu W.
        Serum neurofilament light chain at time of diagnosis is an independent prognostic factor of survival in amyotrophic lateral sclerosis.
        Eur. J. Neurol. 2020; 27: 251-257https://doi.org/10.1111/ene.14063