Analytical validation of a multi-protein, serum-based assay for disease activity assessments in multiple sclerosis

Proximity extension assay-based proteomics studies in neurodegenerative disorders and multiple sclerosis

Neurodegenerative diseases impact the structure and operation of the nervous system, causing progressive and irreparable harm. Efforts for distinguishing neurodegenerative diseases in their early stages are continuing. Despite several biomarkers being identified, there is always search for more accurate and abundant ones. Additionally, it can be difficult to pinpoint the precise neurodegenerative disorder affecting a patient as the symptoms of these conditions frequently overlap. Numerous studies have shown that pathological changes occur years before clinical signs appear. Therefore, it is crucial to discover blood-based biomarkers for neurodegenerative diseases for easier and earlier diagnosis. Proximity extension assay is a unique proteomics method that uses antibodies linked to oligonucleotides for quantifying proteins with real-time PCR. Proximity extension assay can identify even low-quantity proteins using a small volume of specimens with increased sensitivity compared to conventional methods. In this article, we reviewed the employment of proximity extension assay technology to detect biomarkers or protein profiles for several neurodegenerative diseases.

Proteomic signatures of physical, cognitive, and imaging outcomes in multiple sclerosis

BACKGROUND

A quantitative measurement of serum proteome biomarkers that would associate with disease progression endpoints can provide risk stratification for persons with multiple sclerosis (PwMS) and supplement the clinical decision-making process.

MATERIALS AND METHODS

In total, 202 PwMS were enrolled in a longitudinal study with measurements at two time points with an average follow-up time of 5.4 years. Clinical measures included the Expanded Disability Status Scale, Timed 25-foot Walk, 9-Hole Peg, and Symbol Digit Modalities Tests. Subjects underwent magnetic resonance imaging to determine the volumetric measures of the whole brain, gray matter, deep gray matter, and lateral ventricles. Serum samples were analyzed using a custom immunoassay panel on the Olink™ platform, and concentrations of 18 protein biomarkers were measured. Linear mixed-effects models and adjustment for multiple comparisons were performed.

RESULTS

Subjects had a significant 55.6% increase in chemokine ligand 20 (9.7 pg/mL vs. 15.1 pg/mL, p < 0.001) and neurofilament light polypeptide (10.5 pg/mL vs. 11.5 pg/mL, p = 0.003) at the follow-up time point. Additional changes in CUB domain-containing protein 1, Contactin 2, Glial fibrillary acidic protein, Myelin oligodendrocyte glycoprotein, and Osteopontin were noted but did not survive multiple comparison correction. Worse clinical performance in the 9-HPT was associated with neurofilament light polypeptide (p = 0.001). Increases in several biomarker candidates were correlated with greater neurodegenerative changes as measured by different brain volumes.

CONCLUSION

Multiple proteins, selected from a disease activity test that represent diverse biological pathways, are associated with physical, cognitive, and radiographic outcomes. Future studies should determine the utility of multiple protein assays in routine clinical care.

Association between serum multi-protein biomarker profile and real-world disability in multiple sclerosis

Few studies examined blood biomarkers informative of patient-reported outcome (PRO) of disability in people with multiple sclerosis (MS). We examined the associations between serum multi-protein biomarker profiles and patient-reported MS disability. In this cross-sectional study (2017-2020), adults with diagnosis of MS (or precursors) from two independent clinic-based cohorts were divided into a training and test set. For predictors, we examined seven clinical factors (age at sample collection, sex, race/ethnicity, disease subtype, disease duration, disease-modifying therapy [DMT], and time interval between sample collection and closest PRO assessment) and 19 serum protein biomarkers potentially associated with MS disease activity endpoints identified from prior studies. We trained machine learning (ML) models (Least Absolute Shrinkage and Selection Operator regression [LASSO], Random Forest, Extreme Gradient Boosting, Support Vector Machines, stacking ensemble learning, and stacking classification) for predicting Patient Determined Disease Steps (PDDS) score as the primary endpoint and reported model performance using the held-out test set. The study included 431 participants (mean age 49 years, 81% women, 94% non-Hispanic White). For binary PDDS score, combined feature input of routine clinical factors and the 19 proteins consistently outperformed base models (comprising clinical features alone or clinical features plus one single protein at a time) in predicting severe (PDDS ≥ 4) versus mild/moderate (PDDS < 4) disability across multiple machine learning approaches, with LASSO achieving the best area under the curve (AUCPDDS = 0.91) and other metrics. For ordinal PDDS score, LASSO model comprising combined clinical factors and 19 proteins as feature input (R2PDDS = 0.31) again outperformed base models. The two best-performing LASSO models (i.e., binary and ordinal PDDS score) shared six clinical features (age, sex, race/ethnicity, disease subtype, disease duration, DMT efficacy) and nine proteins (cluster of differentiation 6, CUB-domain-containing protein 1, contactin-2, interleukin-12 subunit-beta, neurofilament light chain [NfL], protogenin, serpin family A member 9, tumor necrosis factor superfamily member 13B, versican). By comparison, LASSO models with clinical features plus one single protein at a time as feature input did not select either NfL or glial fibrillary acidic protein (GFAP) as a final feature. Forcing either NfL or GFAP as a single protein feature into models did not improve performance beyond clinical features alone. Stacking classification model using five functional pathways to represent multiple proteins as meta-features implicated those involved in neuroaxonal integrity as significant contributors to predictive performance. Thus, serum multi-protein biomarker profiles improve the prediction of real-world MS disability status beyond clinical profile alone or clinical profile plus single protein biomarker, reaching clinically actionable performance.

Lower levels of Th1 and Th2 cytokines in cerebrospinal fluid (CSF) at the time of initial CSF shunt placement in children are associated with subsequent shunt revision surgeries

OBJECTIVE

We compare cytokine profiles at the time of initial CSF shunt placement between children who required no subsequent shunt revision surgeries and children requiring repeated CSF shunt revision surgeries for CSF shunt failure. We also describe the cytokine profiles across surgical episodes for children who undergo multiple subsequent revision surgeries.

METHODS

This pilot study was nested within an ongoing prospective multicenter study collecting CSF samples and clinical data at the time of CSF shunt surgeries since August 2014. We selected cases where CSF was available for children who underwent an initial CSF shunt placement and had no subsequent shunt revision surgeries during >=24 months of follow-up (n = 7); as well as children who underwent an initial CSF shunt placement and then required repeated CSF shunt revision surgeries (n = 3). Levels of 92 human cytokines were measured using the Olink immunoassay and 41 human cytokines were measured using Luminex based bead array on CSF obtained at the time of each child's initial CSF shunt placement and were displayed in heat maps.

RESULTS

Qualitatively similar profiles for the majority of cytokines were observed among the patients in each group in both Olink and Luminex assays. Lower levels of MCP-3, CASP-8, CD5, CXCL9, CXCL11, eotaxin, IFN-γ, IL-13, IP-10, and OSM at the time of initial surgery were noted in the children who went on to require multiple surgeries. Pro- and anti-inflammatory cytokines were selected a priori and shown across subsequent revision surgeries for the 3 patients. Cytokine patterns differed between patients, but within a given patient pro-inflammatory and anti-inflammatory cytokines acted in a parallel fashion, with the exception of IL-4.

CONCLUSIONS

Heat maps of cytokine levels at the time of initial CSF shunt placement for each child undergoing only a single initial CSF shunt placement and for each child undergoing repeat CSF shunt revision surgeries demonstrated qualitatively similar profiles for the majority of cytokines. Lower levels of MCP-3, CASP-8, CD5, CXCL9, CXCL11, eotaxin, IFN-γ, IL-13, IP-10, and OSM at the time of initial surgery were noted in the children who went on to require multiple surgeries. Better stratification by patient age, etiology, and mechanism of failure is needed to develop a deeper understanding of the mechanism of inflammation in the development of hydrocephalus and response to shunting in children.

Inflammatory and neurodegenerative serum protein biomarkers increase sensitivity to detect disease activity in multiple sclerosis

Background/Objectives

Serum proteomic analysis of deeply-phenotyped samples, biological pathway modeling and network analysis were performed to elucidate the inflammatory and neurodegenerative processes of multiple sclerosis (MS) and identify sensitive biomarkers of MS disease activity (DA).

Methods

Over 1100 serum proteins were evaluated in >600 samples from three MS cohorts to identify biomarkers of clinical and radiographic (gadolinium-enhancing lesions) new MS DA. Protein levels were analyzed and associated with presence of gadolinium-enhancing lesions, clinical relapse status (CRS), and annualized relapse rate (ARR) to create a custom assay panel.

Results

Twenty proteins were associated with increased clinical and radiographic MS DA. Serum neurofilament light chain (NfL) showed the strongest univariate correlation with radiographic and clinical DA measures. Multivariate modeling significantly outperformed univariate NfL to predict gadolinium lesion activity, CRS and ARR.

Discussion

These findings provide insight regarding correlations between inflammatory and neurodegenerative biomarkers and clinical and radiographic MS DA.

Funding

Octave Bioscience, Inc (Menlo Park, CA).

Proteomics and relationship with axonal pathology in multiple sclerosis: 5-year diffusion tensor imaging study

Blood-based biomarkers can be economic and easily accessible tools for monitoring and predicting disease activity in multiple sclerosis. The objective of this study was to determine the predictive value of a multivariate proteomic assay for concurrent and future microstructural/axonal brain pathology in a longitudinal study of a heterogeneous group of people with multiple sclerosis. A proteomic analysis was obtained on serum samples from 202 people with multiple sclerosis (148 relapsing-remitting and 54 progressive) at baseline and 5-year follow-up. The concentration of 21 proteins related to multiple pathways of multiple sclerosis pathophysiology was derived using Proximity Extension Assay on the Olink platform. Patients were imaged on the same 3T MRI scanner at both timepoints. Тhe rate of whole brain, white matter and grey matter atrophy over the 5-year follow-up was determined using the multi-timepoint Structural Image Evaluation, using Normalisation, of Atrophy algorithms. Lesion burden measures were also assessed. The severity of microstructural axonal brain pathology was quantified using diffusion tensor imaging. Fractional anisotropy and mean diffusivity of normal-appearing brain tissue, normal-appearing white matter, grey matter, T2 and T1 lesions were calculated. Age, sex and body mass index-adjusted step-wise regression models were used. Glial fibrillary acidic protein was the most common and highest-ranked proteomic biomarker associated with greater concurrent microstructural central nervous system alterations (P < 0.001). The rate of whole brain atrophy was associated with baseline levels of glial fibrillary acidic protein, protogenin precursor, neurofilament light chain and myelin oligodendrocyte (P < 0.009), whereas grey matter atrophy was associated with higher baseline neurofilament light chain, higher osteopontin and lower protogenin precursor levels (P < 0.016). Higher baseline glial fibrillary acidic protein level was a significant predictor of future severity of the microstructural CNS alterations as measured by normal-appearing brain tissue fractional anisotropy and mean diffusivity (standardized β = -0.397/0.327, P < 0.001), normal-appearing white matter fractional anisotropy (standardized β = -0.466, P < 0.0012), grey matter mean diffusivity (standardized β = 0.346, P < 0.011) and T2 lesion mean diffusivity (standardized β = 0.416, P < 0.001) at the 5-year follow-up. Serum levels of myelin-oligodendrocyte glycoprotein, neurofilament light chain, contactin-2 and osteopontin proteins were additionally and independently associated with worse concomitant and future axonal pathology. Higher glial fibrillary acidic protein levels were associated with future disability progression (Exp(B) = 8.65, P = 0.004). Multiple proteomic biomarkers are independently associated with greater severity of axonal brain pathology as measured by diffusion tensor imaging in multiple sclerosis. Baseline serum glial fibrillary acidic protein levels can predict future disability progression.