Combined analysis of cytokines and neurofilaments improves differentiation and prognostication in ALS

doi:10.21203/rs.3.rs-2126255/v1

Download PDF

Research Article

Combined analysis of cytokines and neurofilaments improves differentiation and prognostication in ALS

https://doi.org/10.21203/rs.3.rs-2126255/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Amyotrophic lateral sclerosis (ALS) is a heterogeneous neurodegenerative syndrome hallmarked by adult-onset degeneration of upper and lower motor neurons and their associated tracts. ALS may be difficult to differentiate from some other neurological disorders termed ALS mimics and to estimate prognosis.

Methods

We retrospectively measured the concentrations of neurofilaments (NFs) and cytokines in cerebrospinal fluid (CSF) and plasma in ALS patients (n = 234) and ALS mimics (n = 44), to assess their association to diagnosis and prognosis in ALS. Two in-house developed enzyme-linked immunosorbent assays were used to measure NF light chain (NFL) and phosphorylated NF heavy chain (pNFH) in CSF. Single-molecule array was used to measure NFL in plasma and cytokines in CSF and plasma.

Results

Significantly higher concentrations of CSF tumor necrosis factor (TNF)-α and lower concentrations of plasma interleukin (IL)-1β were found in bulbar-onset ALS compared with ALS mimics. Plasma IL-6, plasma IL-17A and plasma TNF-α concentrations in ALS patients were significantly negatively correlated with the ALS functional rating scale revised and significantly positively correlated with King’s ALS clinical staging. Furthermore, plasma IL-6 and plasma TNF-α concentrations were significantly negatively correlated with the disease duration and significantly positively correlated with the disease progression rate in ALS patients. In both CSF and plasma in ALS patients, TNF-α was significantly positively correlated with NFL. ALS patients having plasma NFL or plasma IL-6 concentrations equal to or higher than the upper quartile (≥ Q3) showed a shorter disease duration compared with ALS patients having plasma NFL or plasma IL-6 equal to or lower than the median (≤ Q2).

Conclusions

Combined analysis of inflammatory and neurodegenerative biomarkers may be useful in differentiation between ALS and ALS-mimics. Such analysis may also provide useful prognostic information and facilitate stratification of less heterogeneous groups in clinical drug trials.

Amyotrophic lateral sclerosis

ALS mimics

Cytokines

Interleukin

Neurofilaments

Inflammation

Neurodegeneration

SOD1

C9orf72

Diagnostic sensitivity

Single-molecule array

Amyotrophic lateral sclerosis (ALS) is a phenotypically (1), genetically (2) and prognostically (3, 4) heterogeneous syndrome hallmarked by adult-onset progressive degeneration of primarily upper and lower motor neurons and their associated tracts (1), resulting in wasting of skeletal muscles and eventually death from paralysis of the respiratory muscles or pneumonia. In the majority of cases, symptoms appear first in a single limb termed spinal onset ALS, or in the head and neck region termed bulbar-onset ALS (1). Mutations in over 40 genes have been associated with ALS; the two most common being mutations in SOD1 and a hexanucleotide repeat expansion in C9orf72 (C9orf72HRE) (2). A wide variety of conditions can present with symptoms similar to ALS, termed ALS mimics (5, 6). Misdiagnosis in ALS is frequent particularly in the early stages of ALS and some patients are only diagnosed late during the disease course or at a post-mortem autopsy (7). While the median survival time from symptom onset to death ranges between 2–4 years, around 10% of patients with ALS survive more than 10 years (3, 8). The heterogeneity in disease duration is also a major issue when designing clinical drug trials where enrollment of patients early in the disease course is most desirable; however, this interest competes with a higher uncertainty of the diagnosis and prognosis. Previous studies suggest neurofilament (NF) light chain (NFL) and phosphorylated NF heavy chain (pNFH) being useful biomarkers for differentiating ALS from other neurological diseases and for estimating prognosis in ALS (9–19). The role of systemic cytokine signaling and inflammation in ALS is contentious, with mixed results when comparing cytokine concentrations in ALS with controls (20–23). Previous studies suggest elevated levels of cytokines, chemokines and growth factors in ALS patients, including Interleukin (IL)-1β, IL-6, IL-10, IL-17A, interferon (IFN)-γ, tumor necrosis factor (TNF)-α, monocyte chemoattractant protein 1 (MCP-1) and vascular endothelial growth factor (21, 24–27). Patients with multifocal motor neuropathy (MMN) have previously shown significantly higher concentrations of serum TNF-α compared with ALS patients and cases with other non-inflammatory neurological diseases (OND) (28). This observation suggests that cytokines may be useful in differentiating ALS from clinically relevant ALS mimics. Furthermore, ALS patients show extensive intragroup variations in terms of NF estimates and disease duration (19). Several cytokines have been associated with worse survival in ALS and differences in cytokine concentrations have been observed depending on the genotype of the ALS patients (20, 27). Combining the results of cytokines and NFs to differentiate ALS from diagnostic mimics and for stratifying ALS patients into less heterogeneous prognostic groups could potentially be beneficial also for selecting patients for clinical drug trials. In the present study, we investigated the diagnostic and prognostic value of combined analysis of cytokines and NFs in cerebrospinal fluid (CSF) and plasma, also incorporating clinical parameters the patients clinical stage and functional status when the samples were obtained, the disease progression rate and weight alteration since onset of ALS.

Research participants

The study participants consist of patients referred to the Department of Neurology, University Hospital of Umeå, Sweden, to be examined for possible motor neuron disease during 1994–2016. The cohort has previously been described in detail (19). In brief, patients were clinically evaluated in accordance with the European Federation of Neurological Societies (EFNS) diagnostic guidelines for the clinical management of ALS (29, 30). Patients with known medical history of severe head trauma or neoplastic conditions, pregnancy, infection in the central nervous system, tuberculosis, HIV, severe systemic inflammatory disease or daily treatment with anti-inflammatory drugs prior to sampling were excluded. Patients who later committed suicide, who were later treated with invasive ventilation through a tracheostomy, or who participated in a clinical drug trial were also excluded.

Genotype, clinical stage, functional status, disease progression rate and weight alteration in ALS patients

All ALS patients underwent genetic testing for mutations in SOD1, C9orf72 and VAPB as previously described (31, 32). Clinical staging at sampling was performed retrospectively using the King’s ALS clinical staging (KSS) (33, 34) and functional status using the ALS functional rating scale revised (ALSFRS-R) (35). These scales were not available when the first patients were examined. For uniformity in scoring, clinical data for all participants were reassessed by two independent raters. The ALS disease progression rate (ΔFS) in ALS patients was calculated as previously described (36). Weight alteration was assessed either through previous weight measurements prior to and at the time of sampling if available, or by self-reported weight and weight alteration in selected cases where applicable. A negative value for the weight alteration indicates weight loss whereas a positive value indicates weight gain since onset of paresis. Most ALS patients were prescribed medication with riluzole at the time of diagnosis and most eventually received a gastrostomy and non-invasive ventilation through a face-mask.

Collection of CSF and plasma samples

The CSF and plasma samples were collected at the initial clinical diagnostic investigation and stored as previously described (19, 27). In brief, in addition to the CSF collected for diagnostic evaluation, an extra 2–4 mL of CSF was collected by lumbar puncture and stored uncentrifuged at -80°C. Also, paired 30 mL of blood was collected by venipuncture in tubes containing ethylenediaminetetraacetic acid (EDTA) as anticoagulant, centrifuged at 1500 g for 15 minutes at + 4°C and aliquoted into plasma, buffy coat and erythrocytes before long-term storage at -80°C. This procedure for collecting extra CSF and blood samples at the diagnostic evaluation is routine practice in our department: should additional diagnostic tests be necessary after the patient has been discharged from the hospital, instead of calling the patient back to the hospital samples can be retrieved from the freezers and analyzed. Should the samples not be needed for additional diagnostic tests, with approval from the Ethical Committee (see below) such samples may be used for research. The CSF and plasma samples had undergone one freeze-thaw cycle prior to thawing the samples for the analyses of cytokines and NFs. EDTA plasma samples were used to reduce the risk for artificially elevated cytokine concentrations compared with serum samples and heparin plasma samples (37).

Biomarker analysis in CSF and plasma

The results of NFs and cytokines analysis in the ALS patients and ALS mimics cohorts have been reported in two separate previous studies (19, 27). In the present study, we combined the gross laboratory result data from the NF and cytokine studies respectively and reevaluated with additional clinical data. Briefly, all laboratory analyses were performed blind to clinical and genotype data. CSF NFL and CSF pNFH concentrations were measured using two in-house developed enzyme-linked immunosorbent assays (ELISAs) (38, 39), with minor modifications for the CSF pNFH assay. CSF and plasma cytokine concentrations and plasma NFL concentrations were measured using single-molecule array (SIMOA) assays on an HD-1 Analyzer (Quanterix, Billerica, MA, USA), as previously described (27, 40). The cytokines previously assayed in CSF and plasma included IL-1β, IL-6, IL-10, IL-17A, IFN-α, TNF-α and TNF-related apoptosis-inducing ligand (TRAIL) and cytokine concentrations below the lower limit of detection (LLOD), defined as signal below blank + 3 standard deviations (SD), were tied at 0.000 pg/mL as previously reported (27) and included in the statistical analyses. Supplementary Table S1 lists the number of CSF and plasma samples assayed for cytokines in ALS and ALS mimics, and the number of samples with cytokine concentrations below LLOD.

Statistical analysis

Statistical analyses were performed using IBM SPSS Statistics version 26 (International Business Machines Corporation, Armonk, NY, USA), except for the robust regression for statistical adjustments of measurable cytokine concentrations which showed significant correlations to freezer storage time, which were performed in Stata version 15 (StataCorp, College Station, TX, USA). The clinical and genotype characteristics in ALS patients are presented in Table 1, and the biomarker concentrations, CSF mono- and polynuclear cell counts and biomarker ratios in ALS patients and ALS mimics are presented as median and lower quartile (Q1) to upper quartile (Q3) in Table 2. Group comparisons were performed using the Mann–Whitney U test where the asymptotic p-value is reported. Bivariate correlations were investigated using Spearman’s rank-order correlation coefficient (ρ) and a bias-corrected and accelerated bootstrap set at 1000 sample runs was applied to generate 95% confidence interval (CI) for ρ. Correlation tests between NF concentrations, ALSFRS-R, ΔFS and disease duration in ALS patients, and between NF concentrations and freezer storage time for ALS patients and ALS mimics were investigated in our previous study and are therefore not performed in the present study (no significant correlation between NF concentrations and freezer storage time were detected in ALS patients or ALS mimics) (19). CSF IL-1β, CSF IL-10, CSF TNF-α, plasma IL-1β and plasma TRAIL concentrations showed either significant positive or negative correlations with the freezer storage time, whereby regression coefficients from robust regressions were used as conservative correction factors to obtain corrected cytokine concentrations for the detected cytokine levels as previously reported (27). Disease duration was defined as the time from patients’ self-reported symptom onset to death, where the symptom onset was defined as the onset of paresis. Kaplan–Meier survival analyses were performed to investigate differences in disease duration for ALS patients with full survival data being available and excluding presently alive patients. The disease duration from symptom onset to death rather than time of sampling to death was used for Kaplan–Meier analyses to generate more accurate survival analyses. The Breslow (Generalised Wilcoxon) test was used to test statistical significance for Kaplan–Meier analyses with the disease duration reported as median and Q1–Q3. To investigate prognostic subgroups, ALS patients were stratified into groups by I) selected biomarkers using biomarker estimates equal to or lower than the median (≤ Q2) and ≥ Q3 as the cut-offs for the ALS patients, and II) weight alteration using ≤ 10% and > 10% weight loss as the cut-off for ALS patients to generate groups for characterization and pairwise comparisons. Receiver operating characteristic (ROC) analyses were used to estimate the diagnostic performance for selected biomarkers and biomarker ratios, comparing ALS symptom onset groups to ALS mimics and are presented as the area under the ROC curve (AUC) and 95% CI for AUC in the results section, whereas AUC, 95% CI for AUC, Youden’s index highest value (J) for estimating the optimal cut-off, biomarker concentration or biomarker ratio cut-off, sensitivity, specificity, positive likelihood ratio (LR+) and negative likelihood ratio (LR-) are presented in Table 3. Furthermore, prognostic ROC analyses for selected biomarker estimate cut-off groups and weight alteration groups were performed by using disease duration as a continuous variable in the ROC analyses (41), where AUC and 95% CI for AUC are reported for the prognostic ROC analyses. ALS patients with truncal onset of paresis were not included in the comparisons with the ALS mimics due to low sample size in the truncal onset ALS group. Statistical tests with p < 0.05 were considered significant.

Table 1

Clinical and genotype characteristics of ALS patients. Values are presented as frequency (n) and proportion (%) or median and Q1–Q3, except for the values of weight alteration which are presented as range between lowest value (most weight loss) and highest value (weight gain or no known weight loss), and are reported as change in percent. Age at symptom onset, age at sampling, disease duration and age at death are presented in years. KSS stages, ALSFRS-R and ΔFS are presented for the time of sampling. Weight prior to sampling and weight at the time of sampling are presented in kilograms. The clinical characteristics of the ALS patients including genotype groups have been described previously(19).
	All ALS	Spinal onset ALS	Bulbar onset ALS	Truncal onset ALS
Sex: Male / Female	124 (53.0) / 110 (47.0)	86 (58.1) / 62 (41.9)	30 (41.7) / 42 (58.3)	7 (63.6) / 4 (36.4)
Mutation status: No known mutation / SOD1 / C9orf72HRE / VAPB	175 (74.8) / 28 (12.0) / 28 (12.0) / 3 (1.2)	100 (67.6) / 27 (18.2) / 19 (12.8) / 2 (1.4)	64 (88.9) / 0 (0.0) / 7 (9.7) / 1 (1.4)	10 (90.9) / 1 (9.1) / 0 (0.0) / 0 (0.0)
Age at symptom onset	63.95; 55.03–70.95	60.47; 48.55–67.62	69.92; 61.68–75.79	65.52; 62.10–70.31
Age at sampling	65.23; 57.63–72.78	62.57; 51.16–69.31	71.40; 62.32–76.75	66.37; 63.29–70.84
Disease duration	2.99; 1.80–4.63	3.36; 2.30–5.65	2.50; 1.49–3.17	2.31; 1.33–4.07
Age at death	67.14; 60.99–75.15	65.38; 57.37–72.26	73.64; 64.54–78.71	68.07; 66.18–74.07
KSS: 1 / 2 / 3 / 4 / N/A	14 (6.0) / 45 (19.2) / 68 (29.1) / 38 (16.2) / 69 (29.5)	10 (6.8) / 33 (22.3) / 48 (32.4) / 17 (11.5) / 40 (27.0)	4 (5.6) / 10 (13.9) / 18 (25.0) / 15 (20.8) / 25 (34.7)	0 (0.00) / 2 (18.2) / 2 (18.2) / 5 (45.4) / 2 (18.2)
Weight prior to sampling	75.0; 63.0–86.0	77.0; 65.5–87.0	68.0; 60.0–84.0	70.5; 58.5–85.0
Weight at sampling	70.0; 61.0–81.0	75.0; 64.0–82.5	64.0; 58.0–81.0	66.0; 58.5–73.25
Weight alteration	-29.3–4.3	-23.0–4.3	-29.3–3.9	-18.7–0.0
ALSFRS-R	42; 38–45	42; 38–45	43; 39–45	35; 32–44
ΔFS	0.45; 0.23–0.90	0.38; 0.22–0.81	0.64; 0.27–1.08	0.48; 0.32–1.56

Table 2

Concentrations of NFs and cytokines in CSF and plasma, CSF mono- and polynuclear cell counts and biomarker ratios in ALS patients and ALS mimics. Values are presented as median and Q1–Q3. Biomarker concentrations are presented in pg/mL and CSF mono- and polynuclear cell counts are presented in 10E6/L. The concentrations of NFs and cytokines in ALS patients including stratification by genotype(19, 27) and the concentrations of NFs in ALS mimics and the subgroups of the ALS mimics(19) have been previously reported.
		All ALS	Spinal-onset ALS	Bulbar-onset ALS	Truncal-onset ALS	ALS mimics
CSF	NFL	4363; 2813–7135	4291; 2639–6994	4885; 3216–7261	3574; 1858–18596	1205; 758–2234
	pNFH	16955; 9313–25707	16682; 8584–25397	18064; 12487–26165	10043; 7034–14253	3651; 2695–5993
	IL-1β	0.097; 0.049–0.204	0.087; 0.048–0.222	0.113; 0.056–0.206	0.102; 0.028–0.202	0.091; 0.043–0.244
	IL-6	2.356; 1.761–3.147	2.352; 1.757–3.121	2.454; 1.790–3.617	1.938; 1.500–3.729	2.271; 1.778–3.274
	IL-10	0.161; 0.129–0.215	0.158; 0.126–0.208	0.172; 0.132–0.216	0.161; 0.129–0.254	0.168; 0.140–0.214
	IL-17A	0.000; 0.000–0.008	0.000; 0.000–0.005	0.000; 0.000–0.012	0.001; 0.000–0.014	0.000; 0.000–0.007
	IFN-α	0.000; 0.000–0.002	0.000; 0.000–0.001	0.000; 0.000–0.003	0.005; 0.000–N/A	0.000; 0.000–0.006
	TNF-α	0.159; 0.101–0.220	0.158; 0.095–0.217	0.162; 0.103–0.239	0.172; 0.129–0.245	0.134; 0.078–0.174
	TRAIL	0.567; 0.451–0.732	0.567; 0.448–0.701	0.573; 0.458–0.796	0.527; 0.264–0.729	0.601; 0.396–0.757
	Mononuclear cell count	0; 0–1	0; 0–1	0; 0–2	0; 0–1	1; 0–2
	Polynuclear cell count	0; 0–1	0; 0–1	0; 0–1	0; 0–0	0; 0–1
Plasma	NFL	156; 99–236	140; 78–203	201; 115–291	162; 65–422	43; 25–80
	IL-1β	0.176; 0.000–0.288	0.174; 0.000–0.320	0.174; 0.000–0.236	0.225; 0.140–0.498	0.223; 0.066–0.497
	IL-6	2.021; 1.125–3.783	1.798; 1.116–3.289	2.364; 1.149–4.975	2.689; 2.248–6.694	1.951; 1.134–3.323
	IL-10	0.606; 0.452–0.854	0.597; 0.427–0.821	0.596; 0.459–0.977	0.757; 0.540–1.325	0.560; 0.475–0.910
	IL-17A	0.196; 0.097–0.465	0.182; 0.087–0.476	0.203; 0.118–0.417	0.211; 0.142–0.450	0.258; 0.082–0.433
	IFN-α	0.003; 0.000–0.021	0.005; 0.000–0.025	0.002; 0.000–0.015	0.008; 0.000–0.025	0.008; 0.000–0.045
	TNF-α	2.981; 2.347–4.911	2.943; 2.268–5.149	3.164; 2.466–4.357	2.888; 2.371–4.980	3.762; 2.203–5.004
	TRAIL	45.921; 29.970–65.060	45.943; 29.997–64.024	46.337; 30.105–69.352	41.132; 30.547–66.057	46.958; 28.294–67.662
Ratios	Albumin_CSF/albumin_plasma	6.6; 4.8–8.6	6.7; 4.8–8.4	6.6; 4.8–8.8	5.7; 3.9–8.1	7.0; 5.4–8.3
	NFL_CSF/IL-1β_plasma	17465; 7452–32127	15174; 4979–27874	22286; 11478–38216	13851; 8804–66609	3186; 1071–5664
	pNFH_CSF/IL-1β_plasma	63020; 23550–113434	53740; 18149–105277	90307; 45544–120107	38240; 27833–81778	9316; 5653–19616
	NFL_plasma/IL-1β_plasma	601; 229–1174	461; 154–921	1043; 549–1414	507; 189–1303	86; 50–234

Table 3

ROC analyses for spinal-onset ALS or bulbar-onset ALS versus ALS mimics for selected biomarkers and biomarker ratios. For plasma IL-1β, the test direction for the ROC analysis was reversed (performing ROC analyses for ALS mimics versus spinal or bulbar onset ALS). Results for ROC analyses are presented as AUC, 95% CI for AUC, J, biomarker concentration (pg/mL) or biomarker ratio cut-off, sensitivity (%), specificity (%), LR+, LR- and p-value. n₁ number of ALS patients, n₂ number of ALS mimics.
		n₁	n₂	AUC	J	Cut-off	Sensitivity	Specificity	LR+	LR-	p-value
Spinal onset ALS	CSF NFL	148	44	0.798; 0.711–0.885	0.595	2532.5	77.7	81.8	4.27	0.27	< 0.0001
	CSF pNFH	140	42	0.872; 0.800–0.944	0.662	7633.5	82.9	83.3	4.97	0.21	< 0.0001
	Plasma NFL	145	42	0.806; 0.722–0.890	0.524	72.18	78.6	73.8	3.00	0.29	< 0.0001
	CSF TNF-α	146	43	0.585; 0.494–0.676	0.221	0.158	50.0	72.1	1.79	0.69	0.091
	Plasma IL-1β	142	41	0.591; 0.490–0.691	0.187	0.177	65.9	52.8	1.40	0.65	0.078
Bulbar onset ALS	CSF NFL	72	44	0.833; 0.742–0.924	0.721	2549	90.3	81.8	4.97	0.12	< 0.0001
	CSF pNFH	69	42	0.893; 0.819–0.966	0.707	9467.7	82.6	88.1	6.94	0.20	< 0.0001
	Plasma NFL	70	42	0.893; 0.822–0.964	0.762	97.35	92.9	83.3	5.57	0.09	< 0.0001
	CSF TNF-α	69	43	0.628; 0.524–0.732	0.263	0.184	44.9	81.4	2.42	0.68	0.023
	Plasma IL-1β	70	41	0.633; 0.520–0.745	0.294	0.369	36.6	92.9	5.12	0.68	0.020
	NFL_CSF/IL-1β_plasma	47	31	0.879; 0.793–0.965	0.689	6082	91.5	77.4	4.05	0.11	< 0.0001
	pNFH_CSF/IL-1β_plasma	45	29	0.920; 0.857–0.982	0.709	43995	77.8	93.1	11.28	0.24	< 0.0001
	NFL_plasma/IL-1β_plasma	46	31	0.897; 0.818–0.976	0.719	269	91.3	80.6	4.72	0.11	< 0.0001

Study population

The study population consisted of 278 participants. Following standard clinical evaluation, 234 received a diagnosis of ALS and 44 an alternative diagnosis. Table 4 present the stratification of the cohort and the ALS-mimics. The 234 ALS patients were grouped into spinal onset (n = 148), bulbar onset (n = 72) or truncal onset ALS (n = 11). One ALS patient was considered to have frontotemporal dementia onset and in two ALS cases the site of symptom onset could not be determined. A subset of the ALS patients had mutations in SOD1 (n = 28), C9orf72HRE (n = 28) or VAPB (n = 3). One ALS patient carried both a C9orf72HRE and a SOD1 D90A mutation and was included in the C9orf72HRE group. One ALS patient with a SOD1 G93S mutation also had a mutation in VAPB and was included in the SOD1 mutation group. Of the 234 ALS patients, 13 ALS patients were alive at the time of the statistical analysis. Reliable data for estimating disease duration was not available for three ALS patients. Autopsy was performed in 26 cases, confirming a diagnosis of ALS in all. After the retrospective assessment of the clinical parameters in ALS patients, KSS stage in 165 patients, ALSFRS-R and ΔFS in 164 patients, and weight alteration in 163 patients could be estimated, respectively.

Table 4

ALS mimics (n = 44) included in the present study. Values are presented as frequency (n). The ALS mimics and their clinical characteristics have been described in a previous study(19).
	Diagnosis	n
Other motor neuron diseases	Spinobulbar muscular atrophy	3
	Progressive muscular atrophy	7
	Primary lateral sclerosis	2
	Motor neuron disease unspecified	1
Neuropathies & myelopathies	Multifocal motor neuropathy	1
	Multiple sclerosis	2
	Progressive supranuclear palsy	1
	Corticobasal degeneration	1
	Cerebellar atrophy of unknown cause	1
	Benign fasciculation syndrome	1
	Dysphagia of unknown cause	2
	Gait disturbance of unknown cause	1
	Spinal stenosis	4
	Other mono- or polyneuropathies	9
	Myelopathy of unknown cause	1
Myopathies	Polymyositis	1
	Inclusion body myositis	1
	Necrotising myopathy	1
	Limb–girdle muscular dystrophy	1
	Other myopathies	3

Cytokines, NFL and pNFH in CSF

CSF TNF-α was significantly higher in bulbar-onset ALS patients compared with ALS mimics (p = 0.023; Fig. 1A), whereas none of the other cytokines measured in CSF showed any significant difference between bulbar-onset ALS and ALS mimics (p > 0.05). There were no significant differences in CSF cytokine concentrations between spinal-onset ALS and ALS mimics groups (p > 0.05), although a similar but non-significant trend for CSF TNF-α was observed in spinal onset ALS patients which attained the lowest p-value of the cytokines measured in CSF when comparing spinal-onset ALS with ALS mimics (p = 0.091; Fig. 1A). ALS mimics showed a higher CSF mononuclear cell count compared with spinal onset ALS (p = 0.046; Fig. 1B), whereas no difference was observed between ALS mimics and bulbar-onset ALS (p = 0.268; Fig. 1B). No differences in CSF polynuclear cell count were observed between ALS mimics and spinal-onset or bulbar-onset ALS patients (p > 0.05 for both).

Of the cytokines and NFs measured in CSF in ALS patients, CSF TNF-α was correlated positively with CSF NFL (ρ = 0.163; 0.036–0.288, p = 0.013) but not with CSF pNFH (ρ = 0.093; -0.028–0.225, p = 0.172). CSF IFN-α correlated negatively with both CSF NFL (ρ=-0.244; -0.400–-0.084, p = 0.005) and CSF pNFH (ρ=-0.191; -0.347–-0.022, p = 0.031), whereas no other significant correlations between the other cytokines in CSF and CSF NFL or CSF pNFH were observed (p > 0.05). ALS patients’ KSS stages were correlated positively with CSF NFL (ρ = 0.174; 0.023–0.345, p = 0.026) but not with CSF pNFH (ρ = 0.041; -0.126–0.205, p = 0.617). ALS patients’ KSS stages were positively correlated with CSF IL-6 (ρ = 0.214; 0.066–0.348, p = 0.006) and CSF IL-10 (ρ = 0.199; 0.040–0.344, p = 0.015) but not with the remainder of the cytokines measured in CSF (p > 0.05). The ALS patients’ ALSFRS-R was negatively correlated with CSF TNF-α (ρ=-0.193; -0.339–-0.043, p = 0.015) whereas none of the other cytokines measured in CSF showed any significant correlation with ALSFRS-R (p > 0.05). Concentrations of cytokines in the CSF did not correlate with ΔFS (p > 0.05).

Cytokines and NFL in plasma

Bulbar-onset ALS patients showed significantly lower plasma IL-1β levels compared with ALS mimics (p = 0.018; Fig. 1C), whereas no other significant differences were observed for the other plasma cytokines when comparing bulbar-onset ALS patients with ALS mimics (p > 0.05). There were no significant differences in plasma cytokine concentrations between spinal-onset ALS and ALS mimics (p > 0.05); however, a similar but non-significant trend was seen for plasma IL-1β which attained the lowest p-value of the cytokines measured in plasma when comparing spinal-onset ALS with the ALS mimic group (p = 0.073; Fig. 1C).

Of the cytokines measured in plasma, only plasma TNF-α showed positive correlation with plasma NFL (ρ = 0.201; 0.068–0.327, p = 0.003), the other cytokines in plasma did not show any significant correlation with plasma NFL (p > 0.05). ALS patients’ KSS stages correlated positively with plasma NFL (ρ = 0.337; 0.190–0.489, p < 0.0001; Fig. 1D), plasma IL-6 (ρ = 0.273; 0.127–0.405, p < 0.001; Fig. 1E), plasma IL-10 (ρ = 0.176; 0.010–0.333, p = 0.032), plasma IL-17A (ρ = 0.229; 0.077–0.378, p = 0.004) and plasma TNF-α (ρ = 0.175; 0.015–0.326, p = 0.027; Fig. 1F), but not with the other cytokines (p > 0.05). ALS patients’ ALSFRS-R correlated negatively with plasma IL-6 (ρ=-0.362; -0.488–-0.226, p < 0.0001), plasma IL-17A (ρ=-0.232; -0.376–-0.057, p = 0.003) and plasma TNF-α (ρ=-0.239; -0.373–-0.081, p = 0.002). The other cytokines did not correlate with ALSFRS-R (p > 0.05). Furthermore, ΔFS correlated positively with plasma IL-6 (ρ = 0.276; 0.115–0.410, p < 0.001) and plasma TNF-α (ρ = 0.230; 0.085–0.370, p = 0.004), but not with other cytokines in plasma (p > 0.05).

Survival and weight alteration in ALS patients

Disease duration in ALS patients correlated negatively with plasma IL-6 (ρ=-0.186; -0.308–-0.059, p = 0.007) and plasma TNF-α (ρ=-0.204; -0.325–-0.071, p = 0.003) but not with the other cytokines analyzed in plasma (p > 0.05). Similar trends for CSF IL-6 (ρ=-0.122; -0.242–-0.002, p = 0.076) and CSF TNF-α (ρ=-0.116; -0.242–0.012, p = 0.091) were observed, however, none of the correlations between any of the cytokines measured in CSF and disease duration reached statistical significance (p > 0.05). The ALS patient cohort was then stratified into ≤ Q2 and ≥ Q3 biomarker estimate cut-off groups to investigate biomarker levels correlating to disease duration and prognosis, respectively. The distributions of the biomarker concentrations and disease duration for the cut-off groups are visualized in Fig. 2A, B. Kaplan–Meier survival analyses showed significantly longer disease duration in ALS patients for the ≤ Q2 biomarker cut-off groups compared with the ≥ Q3 biomarker cut-off groups for plasma NFL (median disease duration of 3.562; 2.366–6.544 years for the ≤ Q2 cut-off group and 1.837; 1.314–2.960 years for the ≥ Q3 cut-off group (p < 0.0001; Fig. 2D), and plasma IL-6 (median disease duration of 3.305; 2.278–6.007 years for the ≤ Q2 cut-off group and 2.511; 1.714–4.129 years for the ≥ Q3 cut-off group (p = 0.019; Fig. 2E). Kaplan–Meier survival analysis did not show any difference in disease duration in ALS patients for plasma TNF-α ≤ Q2 versus ≥ Q3 cut-off groups (p = 0.122; data not shown). Furthermore, combined biomarker cut-off groups were assessed, comparing differences in disease duration and prognostic performance for ALS patients fulfilling the criteria of having both plasma NFL ≤ Q2 and plasma IL-6 ≤ Q2 or having both plasma NFL ≥ Q3 and plasma IL-6 ≥ Q3. A 3D scatter plot was composed to visualize the distributions of plasma NFL, plasma IL-6 and disease duration for the different combined biomarker cut-off groups in ALS patients (Fig. 2C). Kaplan–Meier survival analysis showed a significantly shorter disease duration for the plasma NFL ≥ Q3 and plasma IL-6 ≥ Q3 combined cut-off group, with a median disease duration of 1.766; 1.333–2.538 years compared with the plasma NFL ≤ Q2 and plasma IL-6 ≤ Q2 combined cut-off group with a median disease duration of 3.562; 2.439–7.321 years (p < 0.0001; Fig. 2F).

Kaplan–Meier survival analyses showed a significant difference in disease duration between KSS stages (p < 0.001 for overall comparisons; Fig. 3A). Kaplan–Meier survival analysis also revealed shorter disease duration for ALS patients with > 10% weight loss at diagnosis, with a disease duration of 1.796; 1.205–3.863 years compared to ALS patients with ≤ 10% weight loss, with a disease duration of 2.858; 1.990–5.005 years (p = 0.020; Fig. 3B). KSS stages correlated negatively with disease duration (ρ=-0.342; -0.515–-0.169, p < 0.0001; Fig. 3D) and ALSFRS-R (ρ=-0.533; -0.629–-0.418, p < 0.0001; Fig. 3E), and positively with ΔFS (ρ = 0.369; 0.232–0.506, p < 0.0001). None of the cytokines measured in CSF or plasma showed any significant correlation with the ALS patients’ weight at sampling (p > 0.05). Only CSF IL-10 and plasma IL-17A showed negative correlations with weight alteration (ρ=-0.177; -0.350–-0.004, p = 0.033, and ρ=-0.212; -0.351–-0.063, p = 0.008, respectively) whereas none of the other cytokines showed any significant correlation with weight changes (p > 0.05). Of the NFs measured, only plasma NFL showed significantly negative correlation with ALS patients’ weight at sampling (ρ=-0.203; -0.352–-0.049, p = 0.010), whereas CSF NFL and CSF pNFH did not (p > 0.05). None of the NFs measured showed any significant correlation with weight alteration in the ALS patients (p > 0.05). ALSFRS-R correlated positively with both the patients’ weight at sampling (ρ = 0.213; 0.049–0.365, p = 0.012) and weight alteration (ρ = 0.313; 0.148–0.461, p < 0.001). The ALS patients’ ΔFS correlated negatively with weight alteration (ρ=-0.214; -0.374–-0.021, p = 0.012), but not with weight at the time of sampling (p > 0.05).

ROC analyses for estimating diagnostic and prognostic performance

ROC analyses for spinal-onset ALS versus ALS mimics for the concentrations of NFs showed the highest AUC for CSF pNFH (AUC: 0.872; 0.800–0.944, p < 0.0001), followed by plasma NFL (AUC: 0.806; 0.722–0.890, p < 0.0001) and CSF NFL (0.798; 0.711–0.885, p < 0.0001). ROC analyses for bulbar-onset ALS patients versus ALS mimics showed the highest AUC for CSF pNFH (AUC: 0.893; 0.819–0.966, p < 0.0001; Fig. 1G) and plasma NFL (AUC: 0.893; 0.822–0.964, p < 0.0001), followed by lower AUC for CSF NFL (AUC: 0.833; 0.742–0.924, p < 0.0001). Since significant differences between bulbar-onset ALS patients and ALS mimics were attained for CSF TNF-α (Fig. 1A) and plasma IL-1β (Fig. 1C), ROC analyses were performed for these biomarkers to investigate their diagnostic performance. ROC analyses for bulbar-onset ALS versus ALS mimics for CSF TNF-α (AUC: 0.628; 0.524–0.732, p = 0.023) and for ALS mimics versus bulbar-onset ALS for plasma IL-1β (AUC: 0.633; 0.520–0.745, p = 0.020; Fig. 1H) showed significant results, whereas ROC analyses for spinal-onset ALS versus ALS mimics for CSF TNF-α (AUC: 0.585; 0.494–0.676; p = 0.091) and for ALS mimics versus spinal-onset ALS for plasma IL-1β (AUC: 0.591; 0.490–0.691, p = 0.078) showed similar trends, however, these were not statistically significant. Ratios of NFs and plasma IL-1β were composed, showing modestly improved diagnostic performance for each ratio for bulbar-onset ALS versus ALS mimics compared with solely using the corresponding NF biomarker (Table 3), with the highest AUC for pNFH_CSF/IL-1β_plasma ratio (AUC: 0.920; 0.857–0.982, p < 0.0001; Fig. 1I), followed by NFL_plasma/IL-1β_plasma ratio (AUC: 0.897; 0.818–0.976, p < 0.0001) and NFL_CSF/IL-1β_plasma ratio (AUC: 0.879; 0.793–0.965, p < 0.0001).

ROC analyses were also performed to investigate prognostic performance in ALS patients for selected biomarker cut-off groups (Fig. 2G, H, I) and weight alteration groups (Fig. 3C). Significant findings for estimating ALS patients’ prognosis were achieved for the disease duration for plasma NFL ≤ Q2 versus ≥ Q3 biomarker cut-off groups (AUC: 0.753; 0.674–0.832, p < 0.0001; Fig. 2G) and plasma IL-6 ≤ Q2 versus ≥ Q3 biomarker cut-off groups (AUC: 0.611; 0.519–0.702, p = 0.023; Fig. 2H). When a combined approach for plasma NFL ≤ Q2 and plasma IL-6 ≤ Q2 versus plasma NFL ≥ Q3 and plasma IL-6 ≥ Q3 biomarker cut-off groups was implemented, ROC analysis indicated modestly higher AUC (AUC: 0.775; 0.653–0.898, p < 0.001; Fig. 2I) for stratifying ALS patients into these prognostic groups. Furthermore, ROC analysis investigating disease duration for weight alteration groups in ALS patients dichotomised into ≤ 10% weight loss versus > 10% weight loss also showed significant prognostic performance (AUC: 0.626; 0.507–0.746, p = 0.032; Fig. 3C).

Blood–brain barrier’s relation to biomarkers, onset type and disease duration in ALS

Albumin_CSF/albumin_plasma ratio correlated positively with CSF IL-10 (ρ = 0.201; 0.051–0.336, p = 0.008) and CSF TNF-α (ρ = 0.225; 0.078–0.372, p = 0.002), but not to the remainder of the cytokines measured in CSF or plasma in ALS patients (p > 0.05). Furthermore, albumin_CSF/albumin_plasma ratio was significantly positively correlated with CSF NFL (ρ = 0.278; 0.128–0.421, p < 0.0001), CSF pNFH (ρ = 0.233; 0.084–0.373, p = 0.001) and plasma NFL (ρ = 0.189; 0.044–0.319, p = 0.009) in ALS patients. No significant difference in albumin_CSF/albumin_plasma ratio was observed between spinal-onset ALS and bulbar-onset ALS patients (p > 0.05) nor did the albumin_CSF/albumin_plasma ratio correlate with disease duration (p > 0.05).

Misdiagnosis and delayed diagnosis are major issues in the clinical management of patients with suspected ALS disease. We here tested the hypothesis that combining the results of analysis for a panel of cytokines with neurofilaments levels at the initial diagnostic work up in patients referred for possible ALS disease would improve the diagnostic process. Analyzing retrospectively samples collected at the initial work-up with clinical data collected during the disease course until death, we find that composite analysis not only improves the differentiation between ALS and ALS-mimics, but also provide relevant prognostic data for use in a clinical setting and in selecting patients for recruitment into a clinical drug trial, where a homogenous study population is essential. Here, we report significantly higher CSF TNF-α (Fig. 1A) and lower plasma IL-1β (Fig. 1C) concentrations in bulbar-onset ALS patients compared with ALS mimics. Similar but non-significant trends were observed for spinal-onset ALS patients for CSF TNF-α (Fig. 1A) and plasma IL-1β (Fig. 1C) when compared with ALS mimics. In a previous study a ratio consisting of pNFH and complement component C3 was proposed to improve the diagnostic properties for ALS compared with only using pNFH (15). We composed ratios of NFs and plasma IL-1β, showing modestly improved AUCs for the ratios in differentiating bulbar-onset ALS patients from the ALS mimics included in the present study. An earlier study found no significant differences in serum IL-1β concentrations when comparing patients with ALS, MMN and OND, but the sample size in each group was small and the ALS patients were not stratified according to the site of symptom onset (28). The ALS mimics included in the present study consisted of a clinical relevant but heterogeneous group encompassing ONDs and disorders with an inflammatory pathological profile. Interestingly, both MMN and ALS patients have previously been dreported having elevated serum IL-1 receptor antagonist compared with ONDs (28). It has been proposed that the elevated levels of IL-1 receptor antagonist observed might be a protective feedback response to limit neuronal injury (28). IL-1 receptor antagonist promotes survival of ventral horn neurons and reduce microglial activation in cell culture (42). However, in a pilot study the IL-1 receptor antagonist anakinra showed no significant effect on ALS disease progression, but the study was small (43). Paradoxically, elevated levels of IL-1β have previously been associated with longer survival in ALS patients (20), which could point towards a more complex role of IL-1β in ALS.

We also found that plasma IL-6 and plasma TNF-α levels correlated negatively with ALS disease duration, whereas similar but non-significant trends were observed for CSF IL-6 and CSF TNF-α. Plasma IL-6 and plasma TNF-α both correlated positively with ΔFS, which further strengthens the association of these biomarkers with the prognosis in ALS. For a deeper clinically relevant analysis, the ALS cohort was stratified into ≤ Q2 and ≥ Q3 biomarker concentration cut-off groups for plasma NFL (Fig. 2A) and plasma IL-6 (Fig. 2B). We could then identify a subset of ALS patients included in the ≥ Q3 cut-off group for both plasma NFL and plasma IL-6 (purple dots in the 3D scatter plot in Fig. 2C), with significantly shorter disease duration for the ≥ Q3 cut-off groups compared to the ≤ Q2 cut-off groups (Fig. D, E, F). These results support the combined analysis of plasma IL-6 and plasma NFL as relevant biomarkers for making more accurate prognostication in ALS.

Previous studies suggest that IL-6 and TNF-α promote reduced mitochondrial membrane potential (44, 45), elevated levels of reactive oxygen species (44, 46, 47), stimulation of lipolysis (44, 48, 49), increased leptin mRNA levels and leptin secretion (50, 51) as well as decreased adenosine triphosphate production (44, 45). A study reported TNF-α also promotes neurotoxic outcome due to mitochondrial failure (45). Furthermore, it has been suggested that IL-6 induce muscle atrophy (52, 53). The association between IL-6, TNF-α and leptin is of particular interest since a leptin-deficient mutant SOD1 ALS mouse model showed slowed disease progression, improved survival, decreased motor neuron loss, amended strength and increased body weight and fat mass (54). Additionally, it has been proposed that TNF-α mediates catabolic activation of the ubiquitin–proteasome pathway (55). A study supporting the association between TNF-α and ALS found reduced motor neuron loss, attenuated weight loss and prolonged survival in an ALS transgenic mouse model following treatment inhibiting TNF-α expression (56). Our study shows significant positive correlations in both CSF and plasma between TNF-α and NFL levels in ALS patients, confirming the results of a previous smaller study (20). Combined, the significant positive correlations in CSF and plasma between NFL and TNF-α, the significant positive correlation between plasma TNF-α and ΔFS, and the significant negative correlation between plasma TNF-α and disease duration found in the present study support that chronically elevated TNF-α enhances the disease progression in ALS.

Interestingly, we found significant correlations between numerous cytokines and ALS patients’ KSS stages, including plasma IL-6 (Fig. 1E) and plasma TNF-α (Fig. 1F). The association between IL-6 and TNF-α mediated reduced mitochondrial membrane potential previously described (44, 45) is of interest since reduced complex I activity and higher mitochondrial content has been reported in mononuclear cells from ALS patients (57). A pilot study of the effects of the monoclonal antibody tocilizumab (a IL-6 receptor antagonist) in ALS patients stratified for their baseline inflammation profile, suggested decreased secretion of cytokines in vitro, including of IL-1β, IL-6, IFN-γ and TNF-α (58). Another study found that tocilizumab normalizes the expression of IL-1 and IL-6 in patients with sporadic ALS (59).

A cellular model exposed to CSF from ALS patients showed enhanced release of IL-6 and TNF-α (22, 26, 60). Altered peripheral monocyte composition, compromised endosome–lysosome trafficking and migration of peripheral myeloid cells into the CNS have previously been observed in ALS patients (26, 61). Furthermore, a transgenic SOD1 ALS mouse model with a microRNA inflammatory profile similar to that observed in ALS patients suggests that activation of inflammatory monocytes with recruitment into the spinal cord correlates with neuronal loss; this could be repressed and survival prolonged after treatment with an anti-Ly6C monoclonal antibody (62). This is also of interest considering that ALS patients’ peripheral monocytes differentiated into M1 macrophages have shown an increased pro-inflammatory response with higher levels of IL-6 and TNF-α compared to controls (63).

In the present study, we also investigated the role of the blood–brain barrier’s relation to cytokines, NFs and disease duration in ALS patients. Though we in the present study found that the albumin_CSF/albumin_plasma ratio correlated with the concentrations of CSF TNF-α and NF, a significant correlation between the albumin_CSF/albumin_plasma ratio and disease duration in ALS patients, was not found. This is in agreement with a previous study (64).

Longitudinal measurements of CSF or plasma samples were not included and is a limitation of this study. This might pose a limitation for the prognostication of ALS patients since analyses of longitudinal plasma samples suggest rising concentrations of plasma IL-6 with disease progression (20). As is common in clinical practice, the here studied ALS patients were in different disease stages when they were first admitted to our department. Staging according to KSS revealed significant correlations between disease stages and several cytokines, including plasma IL-6 (Fig. 3F). The present results suggest increased levels of several cytokines in later stages of the disease course in ALS patients, despite not including longitudinal samples. However, longitudinal samples are needed to firmly assess whether elevation of cytokines precedes the conversion to worse clinical stages in ALS.

A previous smaller study to combine cytokines data to aid in ALS diagnosis and prognostication has been reported (65). Developing an ALS biomarker panel involving neurodegenerative, metabolic, mitochondrial and inflammatory biomarkers in CSF and plasma might potentially further increase the diagnostic performance with higher AUC when comparing ALS to other neurological diseases and potentially aid clinicians in stratifying ALS patients into less heterogeneous prognostic groups. The present study highlights the importance of subgrouping ALS patients in terms of clinical symptom onset, clinical stage and biomarker profile when selecting patients for clinical drug trials.

ALS: Amyotrophic lateral sclerosis; C9orf72HRE: Hexanucleotide repeat expansion in C9orf72; NF: Neurofilament; NFL: Neurofilament light chain; pNFH: Phosphorylated neurofilament heavy chain; IL: Interleukin; IFN: Interferon; TNF: Tumor necrosis factor; MMN: Multifocal motor neuropathy; OND: Other non-inflammatory neurological diseases; CSF: Cerebrospinal fluid; KSS: King’s ALS clinical staging; ALSFRS-R: ALS functional rating scale revised; ΔFS: ALS disease progression rate; EDTA: Ethylenediaminetetraacetic acid; ELISA: Enzyme-linked immunosorbent assay; SIMOA: Single-molecule array; SOD1; superoxide dismutase 1; TRAIL: TNF-related apoptosis-inducing ligand; LLOD: Lower limit of detection; SD: Standard deviation; Q1: Lower quartile; Q3: Upper quartile; ρ: Spearman’s rank-order correlation coefficient; CI: Confidence interval; Q2: Median; ROC: Receiver operating characteristic; AUC: Area under the ROC curve; J: Youden’s index highest value; LR+: Positive likelihood ratio; LR-: Negative likelihood ratio.

Ethics approval and consent to participate

The study was approved by the Medical Research Ethical Committee in Sweden (Forskningsetikkommittén, #1994-135 with amendments in 1998, 2003, 2014, 2017 and 2018). The analyses of cytokines in CSF and plasma were performed in Denmark with the approval of the Regional Ethics Committee for Southern Denmark (S-20150002 G). Written informed consent was obtained from all participants and the study was performed in accordance with the 1964 Declaration of Helsinki and its later amendments. Included in the consent form, the participants gave permission that any research results may be published.

Consent for publication

Not applicable.

Availability of data and materials

Data used and analyzed in the present study will be available from the corresponding author upon reasonable request from other investigators adhering to the European Union General Data Protection Regulation (EU) 2016/679 (GDPR).

Competing interests

AB has received financial compensation from Pharma Industry Publishing AB on matters regarding ALS and biomarkers. MNO is currently employed at H. Lundbeck A/S. KF declares no disclosures relevant to the manuscript. FPC declares no disclosures relevant to the manuscript. AET declares no disclosures relevant to the manuscript.AW declares no disclosures relevant to the manuscript. JSM declares no disclosures relevant to the manuscript. IB declares no disclosures relevant to the manuscript. KB has served as a consultant, at advisory boards, or at data monitoring committees for Abcam, Axon, Biogen, JOMDD/Shimadzu. Julius Clinical, Lilly, MagQu, Novartis, Roche Diagnostics, and Siemens Healthineers, and is a co-founder of Brain Biomarker Solutions in Gothenburg AB (BBS), which is a part of the GU Ventures Incubator Program.HZ has served at scientific advisory boards and/or as a consultant for Abbvie, Alector, Annexon, Apellis, Artery Therapeutics, AZTherapies, CogRx, Denali, Eisai, Nervgen, Novo Nordisk, Pinteon Therapeutics, Red Abbey Labs, reMYND, Passage Bio, Roche, Samumed, Siemens Healthineers, Triplet Therapeutics, and Wave, has given lectures in symposia sponsored by Cellectricon, Fujirebio, Alzecure, Biogen, and Roche, and is a co-founder of Brain Biomarker Solutions in Gothenburg AB (BBS), which is a part of the GU Ventures Incubator Program (outside submitted work). NA declares no disclosures relevant to the manuscript. PMA report grants from the Swedish Research Council, the Swedish Brain Foundation, the Knut and Alice Wallenberg Foundation, the Ulla-Carin Lindquist Foundation; has been/is a consultant for Biogen, Orphazyme A/S, Hoffman la Roche/Genentech, Regeneron and AL-S Pharma on matters regarding ALS genetics and clinical trials and receives or has received financial compensation for these services.

Funding

This research has been generously supported by the Swedish Brain Foundation (grants nr. 2012-0262, 2013-0279, 2016-0303, 2018-0310, 2020-0353), the Swedish Research Council (grants nr. 2009-3548, 2012-3167, 2017-03100), the Knut and Alice Wallenberg Foundation (grants nr. 2012.0091, 2014.0305, 2020.0232), the Ulla-Carin Lindquist Foundation, the Kempe Foundation, the Neuroförbundet Association, Umeå University Insamlingsstiftelsen (223-2808-12, 223-1881-13, 2.1.12-1605-14), the Västerbotten County Council (grant nr 56103-7002829), Swedish Brain Power, King Gustaf V:s and Queen Victoria’s Freemason’s Foundation, the University of Southern Denmark, Region of Southern Denmark and Merchant L. F. Foght’s Foundation. AB was supported by research grants from Västernorrland County Council (LVNFOU939000 and LVNFOU967329). FPC was funded by the MIROCALS project from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 633413. AET was supported by Västerbotten County Council. KB was supported by the Swedish Research Council (#2017-00915), the Alzheimer Drug Discovery Foundation (ADDF), USA (#RDAPB-201809-2016615), the Swedish Alzheimer Foundation (#AF-742881), Hjärnfonden, Sweden (#FO2017-0243), the Swedish state under the agreement between the Swedish government and the County Councils, the ALF-agreement (#ALFGBG-715986), and European Union Joint Program for Neurodegenerative Disorders (JPND2019-466-236). HZ is a Wallenberg Scholar supported by grants from the Swedish Research Council (#2018-02532), the European Research Council (#681712 and #101053962), Swedish State Support for Clinical Research (#ALFGBG-71320), the Alzheimer Drug Discovery Foundation (ADDF), USA (#201809-2016862), the AD Strategic Fund and the Alzheimer's Association (#ADSF-21-831376-C, #ADSF-21-831381-C and #ADSF-21-831377-C), the Olav Thon Foundation, the Erling-Persson Family Foundation, Stiftelsen för Gamla Tjänarinnor, Hjärnfonden, Sweden (#FO2019-0228), the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 860197 (MIRIADE), the European Union Joint Programme – Neurodegenerative Disease Research (JPND2021-00694), and the UK Dementia Research Institute at UCL (UKDRI-1003).

Author contributions

AB: major role in data acquisition; statistical analysis of the data; interpretation of the data; drafted and revised the manuscript for intellectual content. MNO: major role in data acquisition; interpretation of the data, revised the manuscript for intellectual content. KF: conceptualisation and design of the study; major role in data acquisition; interpretation of the data; revised the manuscript for intellectual content. FPC: major role in data acquisition; interpretation of the data; revised the manuscript for intellectual content. AET: statistical analysis of the data; interpretation of the data; revised the manuscript for intellectual content. AW: statistical analysis of the data; interpretation of the data; revised the manuscript for intellectual content. JSM: major role in data acquisition; interpretation of the data; revised the manuscript for intellectual content. IB: major role in data acquisition; interpretation of the data; revised the manuscript for intellectual content. KB: major role in data acquisition; interpretation of the data; revised the manuscript for intellectual content. HZ: major role in data acquisition; interpretation of the data; revised the manuscript for intellectual content. NA: major role in data acquisition; interpretation of the data; revised the manuscript for intellectual content. PMA: principal investigator; funding acquisition; conceptualisation and design of the study; patient enrolment and genotyping; major role in data acquisition; interpretation of the data; drafted and revised the manuscript for intellectual content.

Acknowledgements

The authors are indebted to all the patients who made this study possible. The authors thank Ann-Charloth Nilsson, Helena Alstermark, Eva Jonsson, Anne-Louice Eriksson, Maria Lindberg, Karin Hjertkvist, Eva Bern, Sara Egsgaard, Camilla Davidsen, Dorte Aalund Olsen and the laboratory technicians in the Mölndal laboratory for excellent assistance. We thank Miles Trupp for improving the language and discussions. We also thank the many neurologists who over the years have referred patients to us.

Author details

¹Department of Clinical Sciences, Neurosciences, Umeå University, Umeå, Sweden. ²Departments of Regional Health Research and Molecular Medicine, University of Southern Denmark, Odense, Denmark. ³Department of Neurology, Slagelse Hospital, Slagelse, Denmark. ⁴Department of Clinical Immunology, Odense University Hospital, Odense, Denmark. ⁵Department of Psychiatry and Neurochemistry, University of Gothenburg, Gothenburg, Sweden. ⁶Department of Medical Sciences, Clinical Neurophysiology, Uppsala University, 75185 Uppsala, Sweden. ⁷Department of Medical Sciences, Neurology, Uppsala University, 75185 Uppsala, Sweden. ⁸Department of Chemistry, Umeå University, Umeå, Sweden. ⁹Department of Regional Health Research, University of Southern Denmark, Odense, Denmark. ¹⁰Department of Biochemistry and Immunology, Lillebaelt Hospital, University Hospital of Southern Denmark, Vejle, Denmark. ¹¹Clinical Neurochemistry Laboratory, Sahlgrenska University Hospital, Mölndal, Sweden. ¹²Department of Neurodegenerative Disease, UCL Institute of Neurology, London, United Kingdom. ¹³UK Dementia Research Institute at UCL, London, United Kingdom. ¹⁴Hong Kong Center for Neurodegenerative Diseases, Hong Kong, China. ¹⁵Open Patient Data Explorative Network, Odense University Hospital, Odense, Denmark

Swinnen B, Robberecht W. The phenotypic variability of amyotrophic lateral sclerosis. Nat Rev Neurol. 2014;10(11):661-70.
Zou ZY, Zhou ZR, Che CH, Liu CY, He RL, Huang HP. Genetic epidemiology of amyotrophic lateral sclerosis: a systematic review and meta-analysis. J Neurol Neurosurg Psychiatry. 2017;88(7):540-9.
Chio A, Logroscino G, Hardiman O, Swingler R, Mitchell D, Beghi E, et al. Prognostic factors in ALS: A critical review. Amyotroph Lateral Scler. 2009;10(5-6):310-23.
Westeneng H-J, Debray TPA, Visser AE, van Eijk RPA, Rooney JPK, Calvo A, et al. Prognosis for patients with amyotrophic lateral sclerosis: development and validation of a personalised prediction model. Lancet Neurol. 2018;17(5):423-33.
Traynor BJ, Codd MB, Corr B, Forde C, Frost E, Hardiman O. Amyotrophic lateral sclerosis mimic syndromes: a population-based study. Arch Neurol. 2000;57(1):109-13.
Yedavalli VS, Patil A, Shah P. Amyotrophic Lateral Sclerosis and its Mimics/Variants: A Comprehensive Review. J Clin Imaging Sci. 2018;8:53.
Paganoni S, Macklin EA, Lee A, Murphy A, Chang J, Zipf A, et al. Diagnostic timelines and delays in diagnosing amyotrophic lateral sclerosis (ALS). Amyotroph Lateral Scler Frontotemporal Degener. 2014;15(5-6):453-6.
Testa D, Lovati R, Ferrarini M, Salmoiraghi F, Filippini G. Survival of 793 patients with amyotrophic lateral sclerosis diagnosed over a 28-year period. Amyotroph Lateral Scler Other Motor Neuron Disord. 2004;5(4):208-12.
Reijn TS, Abdo WF, Schelhaas HJ, Verbeek MM. CSF neurofilament protein analysis in the differential diagnosis of ALS. J Neurol. 2009;256(4):615-9.
Zetterberg H, Jacobsson J, Rosengren L, Blennow K, Andersen PM. Cerebrospinal fluid neurofilament light levels in amyotrophic lateral sclerosis: impact of SOD1 genotype. Eur J Neurol. 2007;14(12):1329-33.
Gong ZY, Lv GP, Gao LN, Lu Y, Guo J, Zang DW. Neurofilament Subunit L Levels in the Cerebrospinal Fluid and Serum of Patients with Amyotrophic Lateral Sclerosis. Neurodegener Dis. 2018;18(2-3):165-72.
Tortelli R, Ruggieri M, Cortese R, D'Errico E, Capozzo R, Leo A, et al. Elevated cerebrospinal fluid neurofilament light levels in patients with amyotrophic lateral sclerosis: a possible marker of disease severity and progression. Eur J Neurol. 2012;19(12):1561-7.
Tortelli R, Copetti M, Ruggieri M, Cortese R, Capozzo R, Leo A, et al. Cerebrospinal fluid neurofilament light chain levels: marker of progression to generalized amyotrophic lateral sclerosis. Eur J Neurol. 2015;22(1):215-8.
Gaiottino J, Norgren N, Dobson R, Topping J, Nissim A, Malaspina A, et al. Increased neurofilament light chain blood levels in neurodegenerative neurological diseases. PLoS One. 2013;8(9):e75091.
Ganesalingam J, An J, Shaw CE, Shaw G, Lacomis D, Bowser R. Combination of neurofilament heavy chain and complement C3 as CSF biomarkers for ALS. J Neurochem. 2011;117(3):528-37.
Lu CH, Macdonald-Wallis C, Gray E, Pearce N, Petzold A, Norgren N, et al. Neurofilament light chain: A prognostic biomarker in amyotrophic lateral sclerosis. Neurology. 2015;84(22):2247-57.
McCombe PA, Pfluger C, Singh P, Lim CYH, Airey C, Henderson RD. Serial measurements of phosphorylated neurofilament-heavy in the serum of subjects with amyotrophic lateral sclerosis. J Neurol Sci. 2015;353(1-2):122-9.
Gendron TF, Daughrity LM, Heckman MG, Diehl NN, Wuu J, Miller TM, et al. Phosphorylated neurofilament heavy chain: A biomarker of survival for C9ORF72-associated amyotrophic lateral sclerosis. Ann Neurol. 2017;82(1):139-46.
Behzadi A, Pujol-Calderón F, Tjust AE, Wuolikainen A, Höglund K, Forsberg K, et al. Neurofilaments can differentiate ALS subgroups and ALS from common diagnostic mimics. Sci Rep. 2021;11(1):22128.
Lu C-H, Allen K, Oei F, Leoni E, Kuhle J, Tree T, et al. Systemic inflammatory response and neuromuscular involvement in amyotrophic lateral sclerosis. Neurol Neuroimmunol Neuroinflamm. 2016;3(4):e244.
Tortelli R, Zecca C, Piccininni M, Benmahamed S, Dell'Abate MT, Barulli MR, et al. Plasma Inflammatory Cytokines Are Elevated in ALS. Front Neurol. 2020;11:552295.
Huang F, Zhu Y, Hsiao-Nakamoto J, Tang X, Dugas JC, Moscovitch-Lopatin M, et al. Longitudinal biomarkers in amyotrophic lateral sclerosis. Ann Clin Transl Neurol. 2020;7(7):1103-16.
Ngo ST, Steyn FJ, Huang L, Mantovani S, Pfluger CMM, Woodruff TM, et al. Altered expression of metabolic proteins and adipokines in patients with amyotrophic lateral sclerosis. J Neurol Sci. 2015;357(1-2):22-7.
Hu Y, Cao C, Qin X-Y, Yu Y, Yuan J, Zhao Y, et al. Increased peripheral blood inflammatory cytokine levels in amyotrophic lateral sclerosis: a meta-analysis study. Sci Rep. 2017;7(1):9094.
Fiala M, Chattopadhay M, La Cava A, Tse E, Liu G, Lourenco E, et al. IL-17A is increased in the serum and in spinal cord CD8 and mast cells of ALS patients. J Neuroinflammation. 2010;7:76.
Baron P, Bussini S, Cardin V, Corbo M, Conti G, Galimberti D, et al. Production of monocyte chemoattractant protein‐1 in amyotrophic lateral sclerosis. Muscle Nerve. 2005;32(4):541-4.
Olesen MN, Wuolikainen A, Nilsson AC, Wirenfeldt M, Forsberg K, Madsen JS, et al. Inflammatory profiles relate to survival in subtypes of amyotrophic lateral sclerosis. Neurol Neuroimmunol Neuroinflamm. 2020;7(3):e697.
Furukawa T, Matsui N, Fujita K, Miyashiro A, Nodera H, Izumi Y, et al. Increased proinflammatory cytokines in sera of patients with multifocal motor neuropathy. J Neurol Sci. 2014;346(1-2):75-9.
Andersen PM, Abrahams S, Borasio GD, de Carvalho M, Chio A, Van Damme P, et al. EFNS guidelines on the clinical management of amyotrophic lateral sclerosis (MALS)--revised report of an EFNS task force. Eur J Neurol. 2012;19(3):360-75.
Andersen PM, Borasio GD, Dengler R, Hardiman O, Kollewe K, Leigh PN, et al. EFNS task force on management of amyotrophic lateral sclerosis: guidelines for diagnosing and clinical care of patients and relatives. Eur J Neurol. 2005;12(12):921-38.
Synofzik M, Ronchi D, Keskin I, Basak AN, Wilhelm C, Gobbi C, et al. Mutant superoxide dismutase-1 indistinguishable from wild-type causes ALS. Hum Mol Genet. 2012;21(16):3568-74.
Akimoto C, Volk AE, van Blitterswijk M, Van den Broeck M, Leblond CS, Lumbroso S, et al. A blinded international study on the reliability of genetic testing for GGGGCC-repeat expansions in C9orf72 reveals marked differences in results among 14 laboratories. J Med Genet. 2014;51(6):419-24.
Roche JC, Rojas-Garcia R, Scott KM, Scotton W, Ellis CE, Burman R, et al. A proposed staging system for amyotrophic lateral sclerosis. Brain. 2012;135(3):847-52.
Balendra R, Al Khleifat A, Fang T, Al-Chalabi A. A standard operating procedure for King's ALS clinical staging. Amyotroph Lateral Scler Frontotemporal Degener. 2019;20(3-4):159-64.
Cedarbaum JM, Stambler N, Malta E, Fuller C, Hilt D, Thurmond B, et al. The ALSFRS-R: A revised ALS functional rating scale that incorporates assessments of respiratory function. J Neurol Sci. 1999;169(1-2):13-21.
Kimura F, Fujimura C, Ishida S, Nakajima H, Furutama D, Uehara H, et al. Progression rate of ALSFRS-R at time of diagnosis predicts survival time in ALS. Neurology. 2006;66(2):265-7.
Hennø LT, Storjord E, Christiansen D, Bergseth G, Ludviksen JK, Fure H, et al. Effect of the anticoagulant, storage time and temperature of blood samples on the concentrations of 27 multiplex assayed cytokines – Consequences for defining reference values in healthy humans. Cytokine. 2017;97:86-95.
Gaetani L, Hoglund K, Parnetti L, Pujol-Calderon F, Becker B, Eusebi P, et al. A new enzyme-linked immunosorbent assay for neurofilament light in cerebrospinal fluid: analytical validation and clinical evaluation. Alzheimers Res Ther. 2018;10(1):8.
Pujol-Calderón F, Portelius E, Zetterberg H, Blennow K, Rosengren LE, Höglund K. Neurofilament changes in serum and cerebrospinal fluid after acute ischemic stroke. Neurosci Lett. 2019;698:58-63.
Kuhle J, Barro C, Andreasson U, Derfuss T, Lindberg R, Sandelius Å, et al. 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;54(10):1655-61.
Combescure C, Perneger TV, Weber DC, Daurès J-P, Foucher Y. Prognostic ROC Curves: A Method for Representing the Overall Discriminative Capacity of Binary Markers with Right-Censored Time-to-Event Endpoints. Epidemiology. 2014;25(1):103-9.
Schizas N, Andersson B, Hilborn J, Hailer NP. Interleukin‐1 receptor antagonist promotes survival of ventral horn neurons and suppresses microglial activation in mouse spinal cord slice cultures. J Neurosci Res. 2014;92(11):1457-65.
Maier A, Deigendesch N, Muller K, Weishaupt JH, Krannich A, Rohle R, et al. Interleukin-1 Antagonist Anakinra in Amyotrophic Lateral Sclerosis—A Pilot Study. PLoS One. 2015;10(10):e0139684.
Ji C, Chen X, Gao C, Jiao L, Wang J, Xu G, et al. IL-6 induces lipolysis and mitochondrial dysfunction, but does not affect insulin-mediated glucose transport in 3T3-L1 adipocytes. J Bioenerg Biomembr. 2011;43(4):367-75.
Doll DN, Rellick SL, Barr TL, Ren X, Simpkins JW. Rapid mitochondrial dysfunction mediates TNF‐alpha‐induced neurotoxicity. J Neurochem. 2015;132(4):443-51.
Suematsu N, Tsutsui H, Wen J, Kang D, Ikeuchi M, Ide T, et al. Oxidative stress mediates tumor necrosis factor-α-induced mitochondrial DNA damage and dysfunction in cardiac myocytes. Circulation. 2003;107(10):1418-23.
Kastl L, Sauer SW, Ruppert T, Beissbarth T, Becker MS, Süss D, et al. TNF-α mediates mitochondrial uncoupling and enhances ROS-dependent cell migration via NF-κB activation in liver cells. FEBS Lett. 2014;588(1):175-83.
Zhang HH, Halbleib M, Ahmad F, Manganiello VC, Greenberg AS. Tumor Necrosis Factor-α Stimulates Lipolysis in Differentiated Human Adipocytes Through Activation of Extracellular Signal-Related Kinase and Elevation of Intracellular cAMP. Diabetes. 2002;51(10):2929-35.
Plomgaard P, Fischer CP, Ibfelt T, Pedersen BK, van Hall G. Tumor Necrosis Factor-α Modulates Human in Vivo Lipolysis. J Clin Endocrinol Metab. 2008;93(2):543-9.
Trujillo ME, Sullivan S, Harten I, Schneider SH, Greenberg AS, Fried SK. Interleukin-6 Regulates Human Adipose Tissue Lipid Metabolism and Leptin Production in Vitro. J Clin Endocrinol Metab. 2004;89(11):5577-82.
Sarraf P, Frederich RC, Turner EM, Ma G, Jaskowiak NT, Rivet Iii DJ, et al. Multiple cytokines and acute inflammation raise mouse leptin levels: Potential role in inflammatory anorexia. J Exp Med. 1997;185(1):171-5.
Janssen SPM, Gayan-Ramirez G, Van Den Bergh A, Herijgers P, Maes K, Verbeken E, et al. Interleukin-6 causes myocardial failure and skeletal muscle atrophy in rats. Circulation. 2005;111(8):996-1005.
Haddad F, Zaldivar F, Cooper DM, Adams GR. IL-6-induced skeletal muscle atrophy. J Appl Physiol. 2005;98(3):911-7.
Lim MA, Bence KK, Sandesara I, Andreux P, Auwerx J, Ishibashi J, et al. Genetically altering organismal metabolism by leptin-deficiency benefits a mouse model of amyotrophic lateral sclerosis. Hum Mol Genet. 2014;23(18):4995-5008.
Lin S-Y, Chen W-Y, Lee F-Y, Huang C-J, Sheu WH-H. Activation of ubiquitin-proteasome pathway is involved in skeletal muscle wasting in a rat model with biliary cirrhosis: potential role of TNF-α. Am J Physiol Endocrinol Metab. 2005;288(3):E493-501.
Kiaei M, Petri S, Kipiani K, Gardian G, Choi D-K, Chen J, et al. Thalidomide and Lenalidomide Extend Survival in a Transgenic Mouse Model of Amyotrophic Lateral Sclerosis. J Neurosci. 2006;26(9):2467-73.
Ehinger JK, Morota S, Hansson MJ, Paul G, Elmér E. Mitochondrial dysfunction in blood cells from amyotrophic lateral sclerosis patients. J Neurol. 2015;262(6):1493-503.
Mizwicki MT, Fiala M, Magpantay L, Aziz N, Sayre J, Liu G, et al. Tocilizumab attenuates inflammation in ALS patients through inhibition of IL6 receptor signaling. Am J Neurodegener Dis. 2012;1(3):305-15.
Fiala M, Mizwicki MT, Weitzman R, Magpantay L, Nishimoto N. Tocilizumab infusion therapy normalizes inflammation in sporadic ALS patients. Am J Neurodegener Dis. 2013;2(2):129-39.
Mishra P-S, Dhull DK, Nalini A, Vijayalakshmi K, Sathyaprabha TN, Alladi PA, et al. Astroglia acquires a toxic neuroinflammatory role in response to the cerebrospinal fluid from amyotrophic lateral sclerosis patients. J Neuroinflammation. 2016;13(1):212.
Zondler L, Müller K, Khalaji S, Bliederhäuser C, Ruf WP, Grozdanov V, et al. Peripheral monocytes are functionally altered and invade the CNS in ALS patients. Acta Neuropathol. 2016;132(3):391-411.
Butovsky O, Siddiqui S, Gabriely G, Lanser AJ, Dake B, Murugaiyan G, et al. Modulating inflammatory monocytes with a unique microRNA gene signature ameliorates murine ALS. J Clin Invest. 2012;122(9):3063-87.
Du Y, Zhao W, Thonhoff JR, Wang J, Wen S, Appel SH. Increased activation ability of monocytes from ALS patients. Exp Neurol. 2020;328:113259.
Prell T, Vlad B, Gaur N, Stubendorff B, Grosskreutz J. Blood–Brain Barrier Disruption Is Not Associated With Disease Aggressiveness in Amyotrophic Lateral Sclerosis. Front Neurosci. 2021;15:656456.
Gonzalez-Garza MT, Martinez HR, Cruz-Vega DE, Hernandez-Torre M, Moreno-Cuevas JE. Adipsin, MIP-1b, and IL-8 as CSF Biomarker Panels for ALS Diagnosis. Dis Markers. 2018;2018:3023826.

Competing interest reported. AB has received financial compensation from Pharma Industry Publishing AB on matters regarding ALS and biomarkers. MNO is currently employed at H. Lundbeck A/S. KF declares no disclosures relevant to the manuscript. FPC declares no disclosures relevant to the manuscript. AET declares no disclosures relevant to the manuscript. AW declares no disclosures relevant to the manuscript. JSM declares no disclosures relevant to the manuscript. IB declares no disclosures relevant to the manuscript. KB has served as a consultant, at advisory boards, or at data monitoring committees for Abcam, Axon, Biogen, JOMDD/Shimadzu. Julius Clinical, Lilly, MagQu, Novartis, Roche Diagnostics, and Siemens Healthineers, and is a co-founder of Brain Biomarker Solutions in Gothenburg AB (BBS), which is a part of the GU Ventures Incubator Program. HZ has served at scientific advisory boards and/or as a consultant for Abbvie, Alector, Annexon, Apellis, Artery Therapeutics, AZTherapies, CogRx, Denali, Eisai, Nervgen, Novo Nordisk, Pinteon Therapeutics, Red Abbey Labs, reMYND, Passage Bio, Roche, Samumed, Siemens Healthineers, Triplet Therapeutics, and Wave, has given lectures in symposia sponsored by Cellectricon, Fujirebio, Alzecure, Biogen, and Roche, and is a co-founder of Brain Biomarker Solutions in Gothenburg AB (BBS), which is a part of the GU Ventures Incubator Program (outside submitted work). NA declares no disclosures relevant to the manuscript. PMA report grants from the Swedish Research Council, the Swedish Brain Foundation, the Knut and Alice Wallenberg Foundation, the Ulla-Carin Lindquist Foundation; has been/is a consultant for Biogen, Orphazyme A/S, Hoffman la Roche/Genentech, Regeneron and AL-S Pharma on matters regarding ALS genetics and clinical trials and receives or has received financial compensation for these services.

SupplementaryTableS1CombinedanalysisofcytokinesandneurofilamentsimprovesdifferentiationandprognosticationinALS.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Combined analysis of cytokines and neurofilaments improves differentiation and prognostication in ALS

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

Introduction

Methods

Research participants

Genotype, clinical stage, functional status, disease progression rate and weight alteration in ALS patients

Collection of CSF and plasma samples

Biomarker analysis in CSF and plasma

Statistical analysis

Results

Study population

Cytokines, NFL and pNFH in CSF

Cytokines and NFL in plasma

Survival and weight alteration in ALS patients

ROC analyses for estimating diagnostic and prognostic performance

Blood–brain barrier’s relation to biomarkers, onset type and disease duration in ALS

Discussion

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1